Nitrogen oxides (NOx) emissions play an important role in air quality, the nitrogen cycle, and as precursor for climate gasses. The most important sources of NOx emissions are fossil fuel burning (industry and traffic) and the release from soil.With the inversion algorithm DECSO (Daily Emissions Constrained by Satellite Observations) we derive quantitative NOx emissions on a 10-20 km resolution from TROPOMI (Sentinel 5p) observations of NO2, taking advantage of the fine spatial resolution (5 x 3.5 km) of the TROPOMI instrument. The emissions are split into 3 categories: fossil-fuel, soil and maritime emissions. Detailed results will be shown, including the spatial distribution per emission category and time series of specific cities. Among others we see local effects of the COVID-19 regulations, cleaner technology, but also the increasing use of brown-coal. To assess the quality of our emissions, they are compared to the CAMS emission database and regional emission databases focussing on the Iberian peninsula, and Catalonia in particular.

This work focuses on studying signatures of individual ships in TROPOMI NO2 observations. Information on ships, their location and NOx emissions, are obtained from the Ship Traffic Emission Assessment Model (STEAM). For this work altogether 31 large container ships are selected that operated between Europe and Asia between May-October in 2018 and/or 2019. TROPOMI NO2 data is sampled over the Mediterranean along each ships route (information provided by the STEAM model), allowing a maximum of 15 minutes temporal difference between the satellite observation and the ship location. Each of the matching TROPOMI NO2 scene is analysed using the ships route information for the past 2 hours from the satellite observation. For each container ship multiple matching observations are found where a signature of the ships emissions is visible in the TROPOMI NO2 data. These signatures are seen both under sun glint and non-glint conditions, but under glint the signature is often more clear. In this work alltogether 376 potential plume cases are analysed. The TROPOMI NO2 value in the plume is defined from a mean of the three highest NO2 value pixels that fall within a polygon, delimited by the ships route for the past hour as well as expected plume displacement due to prevailing wind conditions. The information on wind and other meteorological parameters is obtained from ERA-5 model. The plume NO2 values vary mainly between 1.5 and 5*1015 molec./cm2, median value being 2*1015 molec./cm2. This corresponds typically about 1*1015 molec./cm2 enhancement as compared to the local background. The plume values are also compared against the STEAM NOx emission estimates. Comparison show that more clear relation is achieved between the plume NO2 and NOx emission estimate, when the data are grouped, e.g. according to meteorological conditions (wind, stability) or glint/non glint cases. This work is funded by the SCIPPER project (H2020 grant agreement Nr. 814893)

This contribution focuses on an improved TROPOMI tropospheric NO2 research product over Europe. We present an overview of the DLR NO2 retrieval algorithm and a validation with ground-based measurements. Furthermore, the use of TROPOMI tropospheric NO2 columns for air quality purposes in Europe will be illustrated. The DLR NO2 retrieval algorithm for the TROPOMI instrument consists of three steps: the spectral fitting of the slant column, the separation of stratospheric and tropospheric contributions, and the conversion of the slant column to a vertical column using an air mass factor (AMF) calculation. To calculate the NO2 slant columns, a 405-465 nm fitting window is applied in the DOAS fit for consistency with other NO2 retrievals from OMI and TROPOMI. Absorption cross-sections of interfering species and a linear intensity offset correction are applied. A directionally dependent STRatospheric Estimation Algorithm from Mainz (DSTREAM) is developed to correct for the dependency of the stratospheric NO2 on the viewing geometry. For AMF calculation, a geometry-dependent effective Lambertian equivalent reflectivity (GE_LER) obtained directly from TROPOMI measurements with a high spatial resolution is used. Mesoscale-resolution a priori NO2 profiles are obtained from the regional POLYPHEMUS/DLR chemistry transport model with the TNO-MACC emission inventory. Based on the latest TROPOMI operational cloud parameters, a more realistic cloud treatment is provided by a clouds-as-layers (CAL) model, which treats the clouds as uniform layers of water droplets, instead of the clouds-as-reflecting-boundaries (CRB) model, in which clouds are simplified as Lambertian reflectors. Validation of the TROPOMI tropospheric NO2 columns is performed by comparisons with ground-based MAXDOAS measurements. The validation is illustrated for nine European stations with urban/suburban conditions. The improved data shows a similar seasonal variation in the tropospheric NO2 columns as the MAX-DOAS measurements with an average correlation coefficient of 0.78 and an average difference of -2.7×1015 molec/cm2. Finally, we present the use of the TROPOMI tropospheric NO2 research product in the regional POLYPHEMUS and LOTOS-EUROS chemistry transport models to analyse the effect of traffic emission on air quality in Germany.

Accurate and up-to-date greenhouse gas (GHG) emission inventories are essential in developing targeted policies designed to reach net zero targets. Efficiently processing the large volumes of data being produced by satellites allows us to detect changes in anthropogenic emissions of GHGs. We have developed a unique analysis tool, using a convolutional neural network (CNN), to identify plumes of nitrogen dioxide (NO2), a tracer of incomplete combustion, from NO2 column data collected by the TROPOspheric Monitoring Instrument (TROPOMI). Our approach allows us to exploit the growing volume of satellite data available to determine the locations of emission hotspots around the globe on a daily time scale. We train the deep learning model using six thousand 28 x 28-pixel images of TROPOMI data (corresponding to ~266 x 133 km2) to find emission plumes of various shapes and sizes. The model can identify plumes with a success rate of more than 90%. Over our study period (July 2018 to June 2020), we detect over 310,000 individual NO2 plumes, each with a corresponding location and timestamp. We can relate these locations to known emission hotspots such as cities, power stations and oil and gas production with over 9% of the detected plumes located over China. Ship tracks through the Suez Canal, South East Asia and around The Cape of Good Hope also become visible within the processed data. We have attempted to remove the influence of open biomass burning using correlative high-resolution thermal infrared data from the Visible Infrared Imaging Radiometer Suite (VIIRS). We also find persistent NO2 plumes from regions where inventories do not currently include emissions, including mid-Africa and Siberia, demonstrating the potential of this tool to help update inventories. We believe our type of analysis, used in conjunction with other earth observation products, could be used to improve estimates of anthropogenic emissions of greenhouse gases.

Trace gases retrieval algorithms of ground-based, airborne and satellite-based remote sensing instruments commonly assume horizontal homogeneity when computing air mass factors (AMF) with radiative transfer models. However, this assumption is not valid for measurements in the vicinity of sources (e.g. cities and power plants), where trace gases have high spatial variability. To study the 3D radiative transfer effects on trace gas retrievals, we implemented 3D box-AMFs and a module to account for the effect of buildings in the MYSTIC solver of the libRadtran radiative transfer model. We demonstrate the importance of 3D radiative transfer by calculating NO2 slant column densities (SCD) for an airborne imaging spectrometer measuring the synthetic emission plume of a point source and the more complex emissions field of traffic in an urban area with mid-rise buildings. Finally, we apply 3D-box AMFs to NO2 measurements from the airborne APEX imaging spectrometer. Our case studies show non-negligible 3D effects due to the slant geometric optical path given by sun, ground pixel and instrument position and due to the effect of atmospheric scattering. These effects become increasingly important when the NO2 field has high spatial variability. 3D effects distort the image of an emission plume and can result in a significant bias in emissions estimated from uncorrected images. The spatial smoothing also partly explains why NO2 maps retrieved from airborne imagers are smoother than, for example, maps simulated with a city-scale dispersion model. Furthermore, the presence of buildings results in a strong and spatially noisy underestimation of NO2, because buildings shield the geometric light path. We conclude that effects of 3D radiative transfer (and buildings) need to be considered when retrieving trace gases from remote sensing instruments at high spatial resolution. 3D radiative transfer effects are important to analyze and reduce spatial smoothing and systematic biases, and allow to better relate the measured slant columns to the true (3D) trace gas distribution. Considering 3D radiative transfer will also become important for future satellite missions with spatial resolutions at and below the kilometer scale. References: Schwaerzel, M., Emde, C., Brunner, D., Morales, R., Wagner, T., Berne, A., Buchmann, B., and Kuhlmann, G.: Three-dimensional radiativetransfer effects on airborne, satellite and ground-based trace gas remote sensing, Atmos. Meas. Tech., doi: https://doi.org/10.5194/amt-13-4277-2020, 2020. Schwaerzel, M., Brunner, D., Jakub, F., Emde, C., Buchmann, B., Berne, A., and Kuhlmann, G.: Impact of 3D radiative transfer on airborne NO2 imaging remote sensing over cities with buildings, Atmos. Meas. Tech. Discuss., doi: https://doi.org/10.5194/amt-2021-129, 2021.

The divergence (spatial derivative) of the horizontal flux of trace gases directly yields the balance of sources and sinks. If applied to TROPOMI measurements of NO2, combined with wind fields from ECMWF, this allows to derive maps of NOx emissions on high spatial resolution (Beirle et al., 2019, 10.1126/sciadv.aax9800).This method is highly sensitive to point sources like power plants, where spatial gradients in the NO2 flux (and thus the divergence) are particularly high. Using a fully automated algorithm for detecting and quantifying local maxima in the divergence map, a global catalog of NOx point sources has been compiled,listing 451 locations identified as NOx point source such as power plants, cement plants, or metal smelters (Beirle et al., 2021, 10.5194/essd-13-2995-2021).Here we introduce the general approach of the divergence method and present the global catalog of NOx point sources, including an update based on recent TROPOMI NO2 processor versions (1.04 and 2.02 involving improved cloud altitudes).

In many cities the population is exposed to elevated levels of air pollution. Often, the spatial distribution of local air quality throughout urban areas is not well known due to the sparseness of official monitoring networks, or due to the inherent limitations of urban air quality models. Satellite observations (e.g., from Sentinel 5P/TROPOMI) and emerging low-cost sensor technology have the potential to provide complementary information. An integrated interpretation, however, is not straightforward. The ESA-funded CitySatAir project was established to contribute towards solving this issue. The project investigates how satellite data of atmospheric composition can be better exploited for monitoring and mapping urban air quality at scales relevant for human exposure. Focusing particularly on the nitrogen dioxide product provided by the TROPOMI instrument on the Sentinel-5P platform, we investigate different approaches for combining this data with other information such as from models, air quality monitoring stations, and low-cost sensor systems. We use two contrasting study sites, namely Madrid, Spain as an example of a large, highly polluted, and mostly cloud-free city and Oslo, Norway as an example of a smaller city with relatively low pollution levels and ubiquitous cloud cover. For Madrid we developed an urban dispersion model able to calculate both surface concentrations of NO2 at street level and NO2 column concentrations matching the TROPOMI observations. The spatial and temporal emissions of the urban area are described with sectoral activity data, for which relevant emission factors must be assigned. When the model is calibrated against ground measurements, it is well capable of reproducing the spatial plume structures seen from space in individual overpasses. We will also show results of the inverse calculation: using TROPOMI retrievals in single or multiple overpasses to estimate the emission fields of the urban area. Once the emission fields are known, the surface concentrations can be calculated with high resolution. For Oslo, we use the Sentinel-5P NO2 data in conjunction with the urban dispersion model EPISODE to bias-correct the underlying bottom-up emission dataset. The results indicate that, when the model is run with the satellite-corrected emission dataset and validated against air quality monitoring stations, the model error (RMSE) decreases for all stations by up to 20%. The updated model dataset is then used to assimilate observations from monitoring stations and low-cost sensors. In addition, we exploit the synergy of TROPOMI and EPISODE data by deriving surface NO2 data and carrying out geostatistical downscaling to provide a satellite-based surface NO2 dataset at scales relevant for human exposure.

Countries around the world have enforced lockdown measures to prevent the spread of the coronavirus disease 2019 (COVID-19), introducing a temporal change of air pollutants such as nitrogen dioxide (NO2) that are strongly related to transportation, industry, and energy. In this study, the NO2 variation over regions with strong responses to COVID-19 is analysed using datasets from Global Ozone Monitoring Experiment-2 (GOME-2) sensor onboard the EUMETSAT Metop satellites and TROPOspheric Monitoring Instrument (TROPOMI) onboard the EU/ESA Sentinel-5 Precursor satellite. The global GOME-2 and TROPOMI NO2 datasets are generated at the German Aerospace Center (DLR); potential influences of long-term trend and seasonal cycle as well as short-term meteorological variation are taken into account statistically. We apply the long-term dataset from GOME-2 and the high-resolution measurements from TROPOMI to analyse the global changes on the tropospheric NO2 columns; regions with strong social responses to COVID-19 in Asia, Europe, North America, and South America show a strong NO2 reduction of ~30-50% on average due to restriction of social and economic activities, followed by a gradual rebounce with lifted restriction measures.

Crude oil production activities and associated petroleum gas (APG) flaring are responsible for significant air polluting and greenhouse gas (GHG) emissions and have negative effects on the environment and climate. In Russia, one of the world's major oil producers, APG flaring remains a routine practice despite regulatory policies. We present the first analysis of nitrogen oxide and methane emissions over Tas-Yuryakh and Talakan oil fields in Sakha Republic (Eastern Siberia, Russia) using multi-satellite observations. Satellite-based TROPOMI (TROPOspheric Monitoring Instrument) nitrogen dioxide (NO2) mean fields show local NO2 enhancements corresponding to the locations of gas flares detected from Sentinel 2 imagery and VIIRS (Visible Infrared Imaging Radiometer Suite) fire data. We derive the annual nitrogen oxide (NOx = NO2+NO) emissions from TROPOMI NO2 observations using an exponentially-modified Gaussian model. We obtain NOx emissions up to 1.34 mol/s (in 2019) in Tas-Yuryakh, where persistent production APG flaring is detected, and about 0.6 mol/s in Talakan, where oil production is three times larger than in Tas-Yuryakh but gas flaring is employed only occasionally. In 2019 we observe a new flaring site in Tas-Yuryakh from the NO2 mean fields, corresponding to an increase in the environmental fees paid by the companies to the local budgets. Assuming that all NOx emissions are associated with APG flaring, the volume of gas flared for 2019 is estimated at 1.25 ± 0.48 billion cubic metres (bcm) in Tas-Yuryakh and 0.5 ± 0.2 bcm in Talakan. Furthermore, we find a clear methane (CH4) anomaly of about 30 ppb from the TROPOMI XCH4 mean fields near Talakan oil field. We estimate mean CH4 emissions of about 28 tons/h from individual TROPOMI XCH4 plumes using the cross-sectional flux method. CH4 enhancements over other oil and gas fields in the area are also detected. The estimated satellite-based NOx and CH4 emissions are higher than the inventories, which are expected to underestimate the contribution from the oil and gas industry and are generally available with several years of delay. TROPOMI NO2 and CH4 observations demonstrate their capability in identifying emission sources from space with unprecedented detail. The results show how satellite observations can support environmental authorities in monitoring the emissions from the oil and gas industry and the commitment of oil companies in reducing APG flaring.

Tropospheric ozone is an important component of the Earth system. It is a greenhouse gas and an oxidant and can therefore affect climate directly, through its radiative effect, and indirectly, through the oxidation of aerosol precursors and its impact on OH and methane. In this work we use ozone in discrete vertical layers from OMI observations, retrieved by the RAL scheme developed in NCEO (with support from CCI/C3S) and compare it to tropospheric ozone from UKESM1 simulations. This observational dataset provides a more versatile tool than the tropospheric column residual widely employed for model evaluation (e.g. Gaudel et al. 2018, TOAR report). Using the OMI dataset it was possible to unequivocally identify model biases affecting tropospheric ozone and the responsible processes. Recent trends in tropospheric ozone were investigated and the discrepancy between model and observations was linked to differences in lower stratospheric ozone trends. Using OMI height-resolved data and model idealised tracers, we were able to identify stratospheric transport of ozone into the troposphere as the main driver of the dynamical response of North Atlantic ozone to modes of variability such as the Arctic Oscillation.

The distribution of ozone in the atmosphere plays a vital role in climate change, air pollution, and human health. Ozone is unequally distributed in the atmosphere, with ~10% found in the troposphere, acting primarily as a pollutant and greenhouse gas. Whereas the remainder is found in the stratosphere, responsible for filtering out harmful ultraviolet (UV) radiation. Ozone has absorption signatures in different regions of the electromagnetic spectrum, which can be exploited using remote sensing techniques, in order to determine the vertical distribution of Ozone. UV satellite estimates of ozone are able to make high-quality estimates of stratospheric concentrations of Ozone, but tend to have lower performance in the troposphere. Thermal infrared (TIR) measurements exploit line pressure broadening to infer tropospheric ozone but typically do not perform as well in estimating stratospheric ozone. The current and future generations of Earth Observation (EO) satellites enable observations from the UV to the TIR. Large volumes of data are available from current satellite instruments such as OMPS and TropOMI in the UV, and AIRS and CrIS in the TIR, and ever-increasing amounts will be available from future instruments such as Sentinels 4, 5 and TEMPO (UV) and future iterations of CrIS (TIR). Optimal utilization of these measurements requires a framework for simultaneous retrievals. We describe the MUltiSpEctra, MUlti-SpEcies, MUlti-SEnsors (MUSES) retrieval algorithm being developed for the TRopospheric Ozone and its Precursors from Earth System Sounding (TROPESS) framework. This algorithm has a flexible and generic Radiative Transfer model that covers the entire wavelength range from the UV to the TIR and a nonlinear optimal estimation retrieval algorithm. Using MUSES/TROPESS, we show examples of multi-sensor retrievals of ozone from Suomi NPP CrIS and Sentinel 5P (S5P)/TROPOMI. The CrIS retrieval operates in a series of micro-windows in the 950-1317.50 cm-1 range, while we investigate two windows in Band 3 of TROPOMI namely 314-340 and 325-335 nm. Suomi NPP orbits in formation with S5P, meaning both satellites view the same scene within minutes of each other allowing for co-location, direct comparison and for joint retrievals. The joint retrieval of CrIS and TROPOMI combines radiances using the previously described windows, and provides additional information content over and above that of each instrument individually. We provide statistics characterising the retrievals from each instrument independently, as well as the joint-retrievals over a range of global conditions, and make comparisons with independent measurements including MUSES-produced OMI and ozonesondes. We evaluate differences with other satellites using the JPL MOMO-Chem chemical data assimilation system. © 2021 California Institute of Technology. US government sponsorship acknowledged.

A long-term tropospheric ozone time series has been generated for the tropical band (20°S to 20°N) based on convective cloud differential algorithm (CCD). Tropical tropospheric ozone columns were retrieved from several European sensors starting with observations by GOME in 1995 and including data from SCIAMACHY, OMI, GOME-2A and GOME-2B. It has now been extended by DLR with data from GOME-2C and TROPOMI and now encompasses 25 years. The tropospheric ozone retrieval for all data sets is based on the total columns retrieved with the GODFIT algorithm and associated cloud products. There are however some differences between the different tropospheric columns from the different sensors which have to be corrected for. For the CCD time series, we used SCIAMACHY data as reference and fitted an offset and a trend correction to the data of the other sensors. We estimated the trend based on the long-term time series. For the tropics an overall trend of +0.7 DU/decade was found in the data set until 2019, varying locally between -0.5 and 1.8 DU/decade. The second data record combines total ozone columns from TROPOMI with BASCOE stratospheric ozone profiles. BASCOE stratospheric ozone data is constrained by assimilated Aura MLS observation and it is provided with 3-hour time resolutions in NRT. We used the BASCOE NRT data set to calculate the stratospheric ozone columns for every day from April 2018 to December 2020 and subtracted it from the respective NRT total columns observed by TROPOMI. The TROPOMI NRT total ozone product was updated recently including a new surface albedo retrieval algorithm. An internal reanalysis of the NRT data was used to create a consistent tropospheric ozone data set. A comparison to ozone sondes showed a good agreement for most part of the world. For the GEMS validation the TROPOMI total ozone NRT algorithm is applied to selected the GEMS data. Also, the tropospheric ozone column might be retrieved based on the TROPOMI-BASCOE algorithm described above. Both the CCD and the TROPOMI-BASCOE tropospheric ozone data will be presented.

We present an update of global and regional total ozone trends 1995-2020 derived from the GOME-type Total Ozone Essential Climate Variable (GTO-ECV) data record. GTO-ECV combines observations from a series of six nadir-viewing satellite sensors of the GOME-type that measure in the ultraviolet and visible spectral range. The excellent inter-sensor consistency, achieved by the common retrieval algorithm GODFIT_v4, is a good prerequisite for merging the individual datasets. GTO-ECV provides monthly mean total ozone at a remarkable spatial resolution of 1°x1°. It has been developed by DLR and as part of the European Space Agency’s Climate Change Initiative (ESA-CCI) ozone project since 2010. In the framework of ESA-CCI+ it has been further improved and expanded with new satellite sensors including TROPOMI onboard Sentinel-5P. As part of the EU Copernicus Climate Change Service (C3S) ozone project it has been extended in time on a quasi-operational basis. Investigating the long-term evolution of ozone is of major interest due to its role as a key constituent in the Earth’s atmosphere. Although the success of the Montreal Protocol and its subsequent amendments is undisputed, and apparent signs of recovery have been detected, open questions remain, in particular with regard to the regional dependence of the trends and their respective uncertainties, as well as the impact of climate change and possible variations in atmospheric dynamics. We report on the spatial as well as the seasonal distribution of the trends and the correlation with the explanatory variables used in the regression model. Regarding the seasonal variation of the trends we found different patterns depending on the latitude range.

We report on updated trends using different merged total ozone datasets from satellite and ground-based observations for the period from 1979 to 2020. This work is an update from the trends reported in Weber et al. (doi:10.5194/acp-18-2097-2018, 2018) using the same datasets up to 2016. Merged datasets used in this study include NASA MOD v8.7 and NOAA Cohesive Data (COH) v8.7, both based on data from the series of Solar Backscatter UltraViolet (SBUV), SBUV-2 and Ozone Mapping and Profiler Suite (OMPS) satellite instruments (1978–present) as well as the Global Ozone Monitoring Experiment (GOME)-type Total Ozone (GTO-ECV) and GOME-SCIAMACHY-GOME-2 (GSG) merged datasets (1995–present), mainly comprising satellite data from GOME, SCIAMACHY, OMI, GOME-2A, -2B’, and TROPOMI. The fifth dataset consists of the annual mean zonal mean data from ground-based measurements collected at the World Ozone and UV Data Center (WOUDC). Trends were determined by applying a multiple linear regression (MLR) to annual mean zonal mean data. The addition of four more years consolidated the fact that total ozone is indeed on its way of slow recovery in both hemispheres as a result of phasing out ozone depleting substances (ODS) as mandated by the Montreal Protocol. The median near global ozone trend after 1996 was 0.5±0.2(2σ) %/decade, which is in absolute numbers roughly a third of the decreasing rate of 1.4±0.6(2σ) %/decade from 1978 until 1996. The ratio of decline and increase is nearly identical to that of the EESC (stratospheric halogen) rates before and after 1996 which proves indeed the success of the Montreal Protocol. The observed trends are also in very good agreement with the median of 17 chemistry climate model (CCMI) with current ODS and GHG scenarios. The positive ODS related trends in the NH after 1996 are only obtained with a sufficient number of terms in the MLR accounting for dynamical ozone changes (Brewer-Dobson circulation, AO, AAO). A standard MLR (limited to solar, QBO, volcanic, and ENSO) leads to zero trends showing that the small positive ODS related trends are balanced by opposite trends in atmospheric dynamics.

The LOTUS community has recently concluded its revisit of long-term changes in stratospheric ozone in support of the upcoming 2022 WMO Ozone Assessment. In this presentation we summarise the progress made since our assessment in 2018 and we present updated estimates of trends in the vertical distribution of ozone between January 2000 and December 2020. All long-term ozone data records by satellite and ground-based instruments have been extended with four additional years of observations. Most of these data records have improved internal coherence, resulting, e.g., from the homogenisation efforts by ground-based networks (NDACC, WMO GAW, SHADOZ) or from refinements to the bias correction schemes to merge satellite data. Improvements were also made to the multiple linear regression model utilized to analyse all time series, e.g., by including seasonal terms for the predictors or by using the updated GloSSAC aerosol proxy. We present the zonal and vertical structure of trends since 2000 and their coherence between data records and simulations by the Chemistry-Climate Model Initiative (CCMI). Ultimately, results from different data sets are combined to obtain our best estimate of trends and uncertainty across the tropical and mid-latitude stratosphere. Our discussion of magnitude and significance of the trends focuses particularly on the upper and lower stratosphere, and we relate our results to earlier findings. Several latitude-longitude resolved merged ozone satellite records have emerged in recent years, such as ESA’s MEGRIDOP, NOAA’s SWOOSH and NASA’s SBUV-MOD. These were analysed for trends in the same way as the zonal mean data records. We discuss regional structures in recent trends inferred from these new satellite data records and we compare these to trends in ground-based records at selected stations where long-term time series are available from lidar, Umkehr and ozonesondes.

Observations in limb geometry from satellite platforms are very valuable to monitor the stratospheric ozone layer on a global scale, as they provide information with high spatial and temporal coverage and a good vertical resolution. At the University of Bremen, observations from two limb sounders, SCIAMACHY (2002-2012) and OMPS-LP (2012-present), were retrieved using the same radiative transfer model, spectroscopic databases and similar retrieval algorithms. The two data sets were merged, to obtain a consistent time series of ozone global distributions. Because of the short overlap of the two missions, measurements performed by the MLS instrument have been used as a transfer function, to provide a statistically significant bias estimate. Monthly latitude- and longitude-resolved time series of ozone profiles were calculated, exploiting the high spatial resolution of the data sets. We used this merged data set to study long-term ozone changes: a multi-linear regression model was applied over the period 2003-2020, finding positive significant ozone trends between 35 and 45 km at mid-latitudes, with an increasing ozone concentration up to 2-4% per decade. Negative but statistically non-significant changes were found in the lower tropical stratosphere. We noticed vertically consistent patterns in the longitude-resolved trends, particularly at northern mid-latitudes above 30 km and in the tropical lower stratosphere. We then performed simulation runs of the TOMCAT global 3-D chemistry transport model for the same period, in order to compare our trend results from the merged data set with the model simulations and to check the consistency of the detected zonal and longitudinally-resolved patterns. Similarities and differences are discussed, with an attempt to refer the latter to specific atmospheric processes (in particular dynamics). Possible explanations of the zonal asymmetries are introduced and a look at seasonally-resolved trends presented.

The satellite measurements in nadir and limb viewing geometry provide a complementary view of the atmosphere. An effective combination of the limb and nadir measurements can provide a new information about atmospheric composition. In this work, we present tropospheric ozone column datasets that have been created using combination of total ozone column from OMI and TROPOMI with stratospheric ozone column dataset from several available limb-viewing instruments (MLS, OSIRIS, MIPAS, SCIAMACHY, OMPS-LP, GOMOS). We have developed further the methodological aspects of assessment of tropospheric ozone using the residual method using simulations with the chemistry-transport model SILAM. It has been shown that the accurate assessment of ozone in the upper troposphere and the lower stratosphere (UTLS) is of high importance for detecting the ground-level ozone patterns.                 The stratospheric ozone column is derived from a combination of ozone profiles from several satellite instruments in limb-viewing geometry. We developed a method for the data homogenization, which includes the removal of biases and a-posteriori estimation (validation) of random uncertainties, thus making the data from different instruments compatible with each other. The high horizontal and vertical resolution dataset of ozone profiles is created via interpolation of the limb profiles from each day to 1°x1° horizonal grid. A new kriging-type interpolation method, which takes into account data uncertainties and the information about natural ozone variations from the SILAM-adjusted ozone field, has been developed. To mitigate the limited accuracy and coverage of the limb profile data in the UTLS, a smooth transition to the model data is applied below the tropopause. This allows estimation of   stratospheric ozone column with full coverage of the UTLS. The derived ozone profiles are in very good agreement with collocated ozonesonde measurements.                 The residual method was successfully applied to OMI and TROPOMI clear-sky total ozone data in combination with the stratospheric ozone column from the high-resolution limb profile dataset. The resulting tropospheric ozone column is in very good agreement with other satellite data. The global distributions of tropospheric ozone exhibit enhancements associated with the regions of high tropospheric ozone production. The main created datasets are (i) monthly 1°x1° global tropospheric ozone column dataset using OMI and limb instruments, (ii) monthly 1°x1° global tropospheric ozone column dataset using TROPOMI and limb instruments and (iii) daily 1°x1° interpolated stratospheric ozone column from limb instruments. Other datasets, which are created as an intermediate step of creating the tropospheric ozone column data, are: (i) daily 1°x1° clear sky and total ozone column from OMI and TROPOMI (ii) Daily 1°x1° homogenized and interpolated dataset of ozone profiles and (iii) daily 1°x1° dataset of ozone profiles from SILAM simulations with adjustment to satellite data. These datasets can be used in various studies related to ozone distributions, variability and trends, both in the troposphere and the stratosphere.

The TOPAS (Tikhonov regularized Ozone Profile retrievAl with SCIATRAN) algorithm to retrieve vertical ozone profiles from space-borne nadir UV radiance measurements was developed and applied to a limited TROPOMI L1B version 2 spectral dataset. Spectral measurements from 270 to 329 nm (band UV1 and UV2) were used after they were spectrally re-calibrated using comparisons to simulated radiances with collocated ozone profiles from MLS/Aura as input. The retrieved ozone profiles have a vertical resolution varying between 6 and 9 km in the stratosphere. Below 18 km the sensitivity is limited and the vertical resolution is reduced. The TOPAS ozone profiles were validated using collocated stratospheric ozone lidar and ozonesonde measurements. The validation with stratospheric ozone profiles shows very good agreement with a mean bias of within ± 5% in the 18 - 45 km altitude range and a standard deviation of 10%. The comparison with MLS and OMPS ozone profiles confirmed the ground based validation results. In the troposphere, the validation with ozonesonde profiles shows a larger bias with up to +40% between 10 and 15 km and a wider spread of results as well. To increase the information content and the vertical resolution below 30 km, CrIS (on Suomi NPP) measurements in the infrared spectral range between 9.35 and 9.9 μm were included in the TOPAS retrieval. The combination of collocated UV TROPOMI and IR CrIS pixels, which have a time difference of only 3 minutes, improves the vertical resolution in the troposphere (above 3 km) to about 10 km. Validation of the combined ozone profiles with tropospheric lidars and ozonesondes shows improved tropospheric profiles and tropospheric ozone column values. In the lower stratosphere up to 30 km, a comparison with MLS ozone profiles shows a slight improvement over the UV-only retrieval.

The total ozone column (TOC) is retrieved using multiple satellite missions. The spatial resolution of total ozone satellite measurements is quite low (e.g., 7x3.5km for TROPOMI and 13x24km for OMI). In some cases (e.g., close to the ozone hole boundary) it is of importance to have information on total ozone at a higher spatial resolution. In this work we propose the use of multiple optical instruments performing the measurements in the visible spectrum Chappuis bands (500-700nm) for the total ozone column determination. This makes it possible to extend the number of instruments, which can be used for the total ozone determination. In particular, current measurements by MODIS/Aqua&Terra, S-GLI/SCOM-C, VIIRS/Suomi-NPP, MSI/S-2, OLCI/S-3 can be also applied for the TOC determination. The total ozone retrievals using Chappuis ozone bands is based on the well-known fact that the top-of-atmosphere reflectance observed over a highly reflective ground (say, snow, ice, deserts) has a minimum in the visible located around 600nm. This feature is due to the absorption of solar light by the atmospheric ozone. The contribution of both ground and atmospheric light scattering to the TOA does not have extrema in the vicinity of 600nm. Therefore, there is a possibility to remove other atmospheric and underlying surface effects to the TOA reflectance over highly reflective underlying surface and derive the atmospheric transmittance due to the ozone absorption effects, which can be used for the TOC determination. This study is aimed at further improvement of the technique as applied to 10 m MSI/S-2 and 300m OLCI/S-3 observations over snow. We have performed intercomparisons of MSI and OLCI TOC retrievals with TOC derived from ground and other satellite (OMI, TROPOMI, GOME-2) measurements. The TOC retrievals using OLCI/S-3 have been performed over entire Antarctica allowing the generation of TOC at various spatial resolutions including 0.5x0.5 and 1x1 degree resolution.

Part of the European Union’s Copernicus Earth Observation programme, the Sentinel-5 Precursor (S5P) satellite mission is dedicated to global atmospheric composition measurements for the monitoring and study of air quality, climate, and both stratospheric and tropospheric ozone. On board of the S5P early afternoon polar satellite, the imaging spectrometer TROPOMI (TROPOspheric Monitoring Instrument) performs nadir measurements of the Earth’s radiance from the UV-visible to near-infrared and short-wave-infrared spectral ranges, from which atmospheric ozone profile data are retrieved. Developed at the Royal Netherlands Meteorological Institute (KNMI) and based on the optimal estimation method, TROPOMI’s operational ozone profile retrieval algorithm (processor version 2.3.0) has recently been implemented into the S5P Payload Data Ground Segment (PDGS). Calibration of the measured radiance and irradiance spectra has also been improved. A diagnostic dataset (DDS5) of fourteen preselected orbits that together collocate with more than one hundred ground-based correlative measurements has been processed by the PDGS afterwards. This work reports on the initial validation of the TROPOMI DDS5 ozone profile data, whereby results are collected from both the operational S5P Mission Performance Centre/Validation Data Analysis Facility (MPC/VDAF) and the S5P Validation Team (S5PVT) scientific project CHEOPS-5p. Based on the same validation practices as developed for and applied to heritage sensors like GOME-2, OMI and IASI, the validation methodology relies on the analysis of data retrieval diagnostics and on comparisons of TROPOMI data with ground-based measurements. The latter are acquired by ozonesondes, stratospheric ozone lidars, and tropospheric ozone lidars, contributing to the NDACC and/or SHADOZ networks, and to the WMO Global Atmosphere Watch programme. The dependence of TROPOMI’s ozone profile uncertainty on several influence quantities like cloud fraction and measurement parameters like sun and scan angles is examined and discussed. This work concludes with a set of quality indicators, enabling users to verify the fitness-for-purpose of the first operational S5P nadir ozone profile data.

Is there a relationship between economic development and air quality? - Especially during the lockdown in Europe, caused by the COVID-19 pandemic, the anthropogenic impact on air pollution received more and more attention. Nitrogen dioxide (NO2) is one of the prominent trace gases in our atmosphere. Within this research study satellite based tropospheric NO2 column measurements obtained from the ERS-2, ENVISAT, MetOp-A and MetOp-B satellites ranging from 1996 to 2021 over the Po valley in northern Italy are analyzed by using two spectral analysis methods. Since most of the NO2 emissions are caused by anthropogenic processes, the results of the spectral analysis methods were compared with the gross domestic product (GDP) as an economic indicator. Furthermore, natural influences on the tropospheric NO2 concentration such as the annual cycle are taken into account. It is found that two major events occurred within the GDP growth rate and the NO2 variability at the same time. The first significant relation occurred in around 2008, caused by the global financial crisis, and was followed by a second correlation which appeared between 2009 and 2014. The findings for the area of interest indicate that for strong economic events such as the global financial crisis or by decreasing foreign investments, it is possible to observe significantly fewer NO2 pollution at the same time.

The Virtual Alpine Observatory (VAO) is a network of 12 European high altitude research stations, based in the Alps and similar mountain regions. It was formed in 2012 to foster the international cooperation between these stations and to join the forces and experiences of many different research teams in order to address the complex and interdisciplinary scientific challenges in the fields of -        Biosphere -        Pedosphere -        Health -        Hydrosphere -        Cryosphere -        Atmosphere With this integrative approach, the network aims to contribute to a better understanding of environmental processes in the Alpine region. Outcomes and results of the work will not only help to support decision makers by balancing economic, social and environmental interests in a sustainable way but shall also set new standards in terms of developing open hardware and provide information products and data analysis tools tailored to the scientific needs. The data gained within the network will be archived and will be available within a delivering service adjusted with specific requirements and addressing special public but also scientific needs. The information technology backbone of VAO is the Alpine Environmental Data and Analysis Center, AlpEnDAC. Elements such as "computing-on-demand", "operating-on-demand" and "services-on-demand" are available here. Satellite-based measurements also play a special role here. Tools that support a comparison of ground-based and satellite-based measurements are also being developed. The office of VAO is hosted by the Bavarian State Ministry of the Environment and Consumer Protection. VAO is part of the European Alpine Convention as well as the Alpine Strategy of the EU. The European Space Agency, ESA, and the Alpine Convention are official VAO observers.

Machine learning algorithms, and especially deep neural networks, provide universal estimator paradigms to approximate optimal solutions for arbitrary domain-specific problems. On the other hand, environmental-related problems that are a direct result of our rapidly changing climate are, nowadays, of the highest importance. Recently, the adoption of machine learning algorithms for environmental modeling has increased, especially in time series forecasting and computer vision. In this review, we attempt to provide a unified and systematic survey of the current machine learning algorithms used to solve multiple air quality monitoring tasks. We specifically focus on air quality modeling using satellite imagery and sensor device data. Lastly, we propose future directions with neural network modeling and representation learning.

Due to rapid increase in urbanization and industrialization Pakistan is facing severe air quality problems which are causing harmful effects on human health and climate system. Therefore, air pollution has turn out to be an increasingly serious issue in Pakistan. In this study an effort has been made to examine the air quality over five megacities (Lahore, Karachi, Islamabad, Peshawar and Quetta) of Pakistan using Sentinel-5P observations during the period 2018–2021. We used satellite based observational datasets for sulphur dioxide (SO2), formaldehyde (HCHO) and carbon monoxide (CO) to analyze the air quality over Pakistan. The results showed that mean values of SO2, HCHO and CO during the study period were 0.000085 mol/m2, 0.000112 mol/m2 and 0.029455 mol/m2 over Pakistan. The maximum concentrations of SO2, HCHO and CO over Pakistan were found in summer, winter and pre-monsoon seasons, respectively. The highest mean concentrations of SO2, HCHO and CO were found to be 0.000197 mol/m2, 0.000236 mol/m2 and 0.040595 mol/m2 over Lahore, respectively, while the lowest concentrations of air pollutants were observed over Quetta during the study period 2018-2020. The results showed that concentration of SO2 increases to about 19.8% over Lahore while HCHO and CO decreases by 2.1% and 1.8% respectively. The concentration of SO2 increases to about 135.7% and 165.6% over Quetta and Peshawar. Similarly, the concentration of HCHO and COalso increases about 45.3% and 2.1% over Islamabad.

The recent World Health Organization’s global air quality guidelines provide new recommendations to reduce the levels of many air pollutants. For example, nitrogen dioxide (NO2) now has an annual recommended level four times smaller than the previous guideline from 2006. These recommendations apply to ground-level measurements which are available from air quality station networks. Satellites can complement surface measurements thanks to their global coverage, but their observations are not directly comparable to ground-level concentrations. In this work, we estimate ground-level NO2 concentrations from satellite-based TROPOMI (on-board the Sentinel-5 Precursor satellite) NO2 observations using the methods by Lamsal et al. (2008) and Cooper et al. (2020) based on the GEOS-Chem chemical transport model. These methods are applied within the area of Finland and compared to in situ measurements to assess their performance.We find that the method by Cooper et al. (2020), which includes a correction for the vertical distribution of NO2 within the boundary layer, overall performs better than the method by Lamsal et al. (2008) with no such correction. When considering 2018–2019 averages, the Cooper et al. (2020) method has a correlation of 0.69 and slope of 0.47 with respect to in situ measurements compared to 0.73 and 0.12, respectively, for the method by Lamsal et al. (2008). There is no clear dependence on the type of air quality station (urban/suburban/rural), but both methods underestimate the NO2 concentrations near highly urban stations in Helsinki. This is expected, as the high vertical and horizontal concentration gradients at such urban canyons are particularly difficult to capture from space, even with TROPOMI’s relatively high spatial resolution.The correction included in the method by Cooper et al. (2020) was calculated based on a sensitivity test across all air quality stations in Finland. The threshold values obtained for the correction are significantly lower than those obtained in the original study for the United States, which correspond to the lower mean concentrations of NO2 measured in Finland. This suggests that the approach is applicable in different regions and mean concentration levels, but that the sensitivity test has to be adapted to the specific area of study.Our results show that TROPOMI satellite observations can be used to estimate surface NO2 concentrations in Finland, and the results are in good agreement with the in situ surface measurements. This can be of interest for national environmental authorities such as the Finnish Ministry of the Environment, which has already incorporated satellite measurements in its latest air quality report to the EU.

The main objective of ESA’s EO4Alps initiative is to foster and establish application services that demonstrate the benefit of EO-derived information for Alpine environmental monitoring, assessment and planning. The application domain “air quality and health” was proposed by ESA as one of four themes that focus on urgent environmental topics for the Alpine region while showing already the potential for operational EO services. Environmental stress by air pollutants and climate belong to the most discussed risk factors for human health. The ESA project AlpAirEO will provide timely information on the aggregated health risk by air pollution and thermal stress covering the Alpine area. This information will be based on earth observation, station network, model and population data. The service interface will be developed following the needs of stakeholders from the health community. For exposure-based risk calculation, we combine environmental data with data on population density and urbanization. Health related monitoring of air pollution asks for high spatial resolution currently not provided on the EU level. We therefore develop a model-based downscaling approach for Copernicus atmospheric data. For verification, medical statistics on the EU and national level will be applied. We present first results of an area-wide assessment of air pollution and the associated health risk in the Alpine space. Preliminary analyses show that the population weighted health burden for main cities exhibits no clear trend albeit intensifying efforts and control measures by EU and local authorities. This underlines the need for consistently monitoring health risk by air pollution in the Alpine region with good areal coverage and high spatial resolution.

The processes that determine Air Quality (AQ) cover a wide range of scales, from point-like emissions to intercontinental transport, and so do the regulations established by public authorities to manage AQ in their area of responsibility. AQ monitoring has hitherto been relying mostly on in-situ measurements of surface concentration, with geographical gaps between observations filled in with numerical modelling. A new generation of satellite sounders on sun-synchronous Low Earth Orbits (LEO) – like the Copernicus Sentinel-5(p) series started in 2017 – performs now daily global mapping of air pollution down to the 3-km scale. Soon this daily global mapping will be complemented with geostationary instruments (GEO, e.g. Sentinel-4) observing the diurnal cycle of pollutants, although over the limited geographical area accessible from a geostationary viewpoint. To realize the full potential of the constellation of LEO and GEO satellites being assembled, and to make their observations fit-for-purpose for air quality applications at the different scales, several challenges need to be addressed, among others: Enhance to sub-city scales the resolution of satellite data, typically from 3-4 km down to 1 km, to make them better suited for the monitoring of e.g. the impact of the Low Emission Zones enforced in several Belgian cities, Determine the non-trivial relation between the column amount of the pollutant measured by a satellite and the surface concentrations measured by in-situ networks, Determine how the different LEO and GEO vantage points lead to a different perception of atmospheric and surface details and how we can benefit from - or correct for - these differences. We present here the developments on these three challenges taking place in the dedicated Belgian federal research project LEGO-BEL-AQ (2020-2023, https://lego-bel-aq.aeronomie.be) funded by BELSPO, which is focused on AQ in Belgium. In response to the first challenge, we demonstrate that a combination of temporal aggregation, careful data selection, and horizontal oversampling can produce a meaningful increase in horizontal resolution in S5P tropospheric NO2 column maps, revealing policy-relevant features in the NO2 distribution over Belgium’s major cities.   The response to the second challenge involves the confrontation between our high-resolution S5P NO2 maps and the in-situ observations (and derived model-based high-resolution gridded surface concentrations) procured by the Belgian authorities. We also already touch upon the project objectives regarding the 3rd challenge, highlighting some key differences between LEO and GEO retrievals of atmospheric trace gases, in particular those related to the oblique viewing geometry inherent to GEO observations of mid- and high latitudes.      Acknowledgements This work has been supported by the BELSPO BRAIN-be 2.0 project LEGO-BEL-AQ (https://lego-bel-aq.aeronomie.be/index.php)

Crowd sourced air pollution projects usually focus on the creation of sensor networks by engaging with citizens together with makerspaces, and through training events. Yet, democratized science must go beyond collecting information, it must enable citizens to take scientifically grounded actions that effectively address their problems, needs, and concerns. In the recently finalized CSEOL project Sentinel Citizen, crowed sourced air pollution data were transformed into capacities by rising awareness of atmospheric commons and by enhancing the actionability of citizen science data. The high-density air pollution network of the citizen science initiative Hollandse Luchten around Amsterdam was combined with remote-sensing data to form observation clusters of air pollution concentration. Based on these clusters we applied a conceptually simple statistical procedure in combination with CAMS data (Copernicus Atmospheric Monitoring Service) to create localized air pollution forecasts in and around Amsterdam. Our presentation will summarize the Atmospheric Commons recommendations formulated in a way that propose very concrete actions to each identified stakeholder. Furthermore, we will present the open and freely accessible methods, procedures and data used to develop the air pollution forecast prototype. CSEOL is an initiative funded by the European Space Agency (ESA) that provides a virtual innovation incubator, i.e. a reliable, facilitating environment to stimulate the emergence of innovative projects that engage citizens, researchers and technology developers in work around ESA’s Earth observation data and which can contribute to empowering citizens, improving decision-making and addressing complex societal problems that can be addressed by using satellite data and citizen science.

Detailed examination of the impact of modern space launches on the Earth’s atmosphere is urgently needed as investment in the space industry booms, people increasingly rely on satellite technologies and as an era of regular space tourism approaches. We add an inventory of pollutant emissions from rocket launches and re-entries of both reusable spacecraft and debris during 2019, to the 3D global chemical transport model GEOS-Chem. We find that chlorine, emitted by solid rocket motors, is the most deleterious to stratospheric ozone (O3), in agreement with previous studies. In addition, a significant proportion of stratospheric O3 loss (33 %, compared to 48 % for chlorine) is caused by nitrogen oxides (NOx), particularly NOx emitted from burn-up of objects re-entering the atmosphere. Annual average depletion of the total stratospheric column is small (< 0.02 %), both in simulations using the 2019 emissions inventory and considering ten years of emissions growth at the rate of launch increase in the last decade, 5.6 % a-1. However, our results pinpoint the region of highest O3 depletion as the upper stratosphere (around 5 hPa, 40 km altitude). At 5 hPa, northern hemisphere (NH) springtime depletion reaches 0.16 % after ten years of emissions growth and accumulation. In a separate simulation modelling the potential impact of regular space tourism flights, we estimate 0.28 % decadal O3 loss at 5 hPa in the NH spring. These values are potentially significant enough to be offsetting the recently observed recovery in upper stratospheric O3 levels due to regulation of ozone depleting substances by the Montreal Protocol.

The special viewing mode of Brewer and Dobson spectrophotometers called Umkehr is used for the retrieval of high quality and high vertical resolution ozone profiles that cover both troposphere, UTLS and the stratosphere. Umkehr measurements can be performed during twilight hours, at high solar zenigh angles (ranging between 60° and 90°) and report ozone in 16 layers extending from the surface to the top of the atmosphere, with each layer being ∼5 km thick. Nevertheless, it is recommended that the profiles are analysed in 8 independent layers, consisting of two layers in the lower atmosphere (0+1, 2+3), five layers in the stratosphere (4, 5, 6, 7, 8) and a broad top layer, combining all layers above 8. The ozone profiles retrieved in this mode can be used for the validation of both satellite IR instrumentation, that cover from the UTLS upwards, as well as UV instrumentation, that also cover the troposphere albeit with a coarser vertical resolution. Within the European Space Agency IDEAS+ QA4EO framework, new efforts are currently underway to improve the operational Umkehr analysis algorithm and thus to provide a more robust and unique validation dataset for ozone profile observations by space-born sensors, such as S5P/TROPOMI, as well as the GOME2 and IASI instruments on the Metop platforms. In this work, we present the first updated Umkehr profiles from Brewer and Dobson spectrophotometers in a number of selected statons. Τhe Umkehr measurements from four Brewer instruments (3 type MKIII and 1 type MKII) operating at Madrid, Spain; Hradec Kralove, Czech Republic; Warsow, Poland and Thessaloniki, Greece were studied for the period 2017 – 2020. The Umkehr measurements from five Dobson stations (Boulder, USA; MLO, USA; Lauder, New Zealand; OHP, France and Arosa, Switzerland) have been optimized and are also presented. Optimized Umkehr records are compared against NASA SBUV v8.6 aggregated overpass records for verification of instrumental bias corrections, as well as the operational EUMETSAT GOME/Metop v2.0 ozone profile products. The overall agreement between Brewer Umkehr profiles and SBUV records is quite satisfactory (±5%), for all layers and stations, with the highest variability observed in the troposphere (layers 0+1, between ~0 and 10 km). The best agreement is found for layers 4 and 5 (between ~20 and 30 km), where the bulk of the ozone absorption occurs. Dobson Umkehr retrievals shows a similar behavior compared to SBUV records also with differences between ±5% for all layers and stations. The optimization procedure has significantly improved the Dobson retrievals as revealed from the comparison with SBUV data. The optimized, quality controlled and fully characterized Umkehr dataset is anticipated to be an added value for the validation of the expected TROPOMI/S5P ozone profile product.

We present a comprehensive study integrating satellite observations of ozone pollution, in situ measurements and chemistry transport model simulations for quantifying the role of anthropogenic emission reductions during the COVID-19 lockdown in spring 2020 over Europe. Satellite observations are derived from the IASI+GOME2 multispectral synergism, which provides particularly enhanced sensitivity to near-surface ozone pollution. These observations are first analysed in terms of differences between the average on 1-15 April 2020, when the strictest lockdown restrictions took place, and the same period in 2019. They show clear enhancements of near-surface ozone in Central Europe and Northern Italy, and some other hotspots, which are typically characterized by VOC-limited chemical regimes. An overall reduction of ozone is observed elsewhere, where ozone chemistry is limited by the abundance of NOx. The spatial distribution of positive and negative ozone concentration anomalies observed from space is in relatively good quantitative agreement with surface in situ measurements over the continent (a correlation coefficient of 0.55, a root-mean-squared difference of 11 ppb and the same standard deviation and range of variability). An average bias of ∼8 ppb between the two observational datasets is remarked, which can partly be explained by the fact the satellite approach retrieves partial columns of ozone with a peak sensitivity above the surface (near 2 km of altitude). For assessing the impact of the reduction of anthropogenic emissions during the lockdown, we adjust the satellite and in situ surface observations for withdrawing the influence of meteorological conditions in 2020 and 2019. This observational estimate of the influence of lockdown emission reduction is consistent for both datasets. They both show lockdown-associated ozone enhancements in hotspots over Central Europe and Northern Italy, with a reduced amplitude with respect to the total changes observed between the two years, and an overall reduction elsewhere over Europe and the ocean. Satellite observations additionally highlight the ozone anomalies in the regions remote from in situ sensors, an enhancement over the Mediterranean likely associated with maritime traffic emissions and a marked large-scale reduction of ozone elsewhere over ocean (particularly over the North Sea), in consistency with previous assessments done with ozonesondes measurements in the free troposphere. These observational assessments are compared with model-only estimations, using the CHIMERE chemistry transport model. For analysing the uncertainty of the model estimates, we perform two sets of simulations with different setups, differing in the emission inventories, their modifications to account for changes in anthropogenic activities during the lockdown and the meteorological fields. Whereas a general qualitative consistency of positive and negative ozone anomalies is remarked between all model and observational estimates, significant changes are seen in their amplitudes. Models underestimate the range of variability of the ozone changes by at least a factor 2 with respect to the two observational data sets, both for enhancements and decreases of ozone, while the large-scale ozone decrease is not simulated.

Elevated concentrations of tropospheric ozone are harmful to ecosystems and human health, in the short term as a pollutant and in the long term as a climate forcer. Global observing systems are faced with the challenge of accurately capturing the considerable spatio-temporal variability in tropospheric ozone at multiple scales resulting from the interplay between dynamical, chemical and radiative processes. Many monitoring studies and climate applications rely on decades-long satellite Climate Data Records (CDR), which combine observations by different sensors in space. The resulting long-term stability and spatial homogeneity of such CDRs depend on our ability to properly account for the differences between the input data records. We present a complete characterisation of the TROPOMI tropospheric ozone data record, providing clues to properly include it in the GOME-type tropospheric ozone CDR (Heue et al., 2016). TROPOMI is a nadir-viewing UV-visible spectrometer launched into space on Sentinel-5 Precursor in October 2017. Since then, it provides daily global coverage at a much finer spatial resolution than its predecessors. The Convective Cloud Differential technique (CCD) is applied to derive three-day running mean ozone columns between surface and 270 hPa over the tropical belt. TROPOMI is characterized primarily by analysing comparisons to SHADOZ ozonesonde and other, currently operating GOME-type sounders (EOS-Aura OMI, Metop-B GOME-2). We will show that TROPOMI bias varies somewhat with reference instrument, but generally remains less than about 2 DU. We find signs of a weak latitudinal pattern and a moderate seasonal pattern in the mean differences, again, depending on the reference instrument. We recommend that these patterns are considered in the merging scheme to avoid artificial changes in the CDR when TROPOMI data are being introduced. The validity of reported random uncertainties is assessed using an analysis of co-located triplets, a technique that is more powerful than the analysis of pairwise measurements. Finally, our verification of the presence of known geophysical structures and cycles confirms that TROPOMI data meet the needs of the atmospheric research community. In particular, we illustrate the resolving power and precision of TROPOMI in capturing the zonal wave-one, the seasonal cycle, biomass burning episodes and intraseasonal variability related to the Madden-Julian Oscillation. Ultimately, this work paves the way for a more comprehensive evaluation of most tropospheric ozone data records contributing to several working groups in the ongoing TOAR-II tropospheric ozone assessement (https://igacproject.org/activities/TOAR/TOAR-II).

Atmospheric ozone plays a key role in air quality and the radiation budget of the Earth, both directly and through its chemical influence on other trace gases. Assessments of the atmospheric ozone distribution and associated climate change therefore demand accurate vertically resolved ozone observations with both stratospheric and tropospheric sensitivity, both on the global and regional scales, and both in the long term and daily to monthly. Such observations have been acquired by two series of European nadir-viewing ozone profilers, namely the scattered-light UV-visible spectrometers of the GOME family, initiated in 1995 (GOME, SCIAMACHY, OMI, GOME-2A/B/C, TROPOMI, and the upcoming Sentinel-5 UVNS series), and the thermal infrared emission sounders of the IASI type, launched regularly since 2006 (IASI on Metop-A/B/C platforms and the upcoming IASI-NG on Metop-SG). Several Level-2 nadir ozone profile datasets have been retrieved from these satellite-based measurements and are now being improved and homogenized in the context of the follow-up project of ESA’s Climate Change Initiative (Ozone_cci+). To verify their fitness-for-purpose, the resulting ozone datasets undergo a harmonized and comprehensive quality assessment (QA), including: (a) detailed identification of their information content and geographical, vertical and temporal representativeness; (b) quantification of their bias, noise and drift, and their dependences on major influence quantities; and (c) assessment of the mutual consistency of data from different sounders. For this purpose, we have applied to the latest nadir ozone profile products within Ozone_cci+ the versatile QA/validation system Multi-TASTE, which has been developed in the context of several heritage projects of ESA, EUMETSAT, and the European Commission. This work reports on harmonized data content studies, information content studies, consistency checks, and ground-based validation for all these ozone climate data records combined, serving as a testbed for similar QA within the CEOS activity on “Tropospheric ozone dataset validation and harmonization” (VC-20-01) in preparation of the second phase of the Tropospheric Ozone Assessment Report (TOAR). The ground-based correlative measurements that are thereby used as a reference are obtained from the Network for the Detection of Atmospheric Composition Change (NDACC), NASA’s Southern Hemisphere Additional Ozonesonde programme (SHADOZ), and other ozonesonde and lidar stations contributing to WMO’s Global Atmosphere Watch (GAW). Dependence of the data quality on major influence quantities is discussed, resulting in data screening suggestions to users.

ALTIUS is a micro-satellite mission whose main objective is to monitor the distribution of stratospheric ozone in the Earth's atmosphere. A typical orbit consists of the following measurement sequence: sunrise occultation, limb scattering observations in the dayside, sunset occultation and, in the nightside, stellar, planetary and lunar occultations. The sequence of stars, moon and planets that will be observed during each night-part of the orbit should be planned to ensure the best spatial coverage and to minimize the time it takes for the satellite to turn to a new target. Here we present the algorithm which is one of the candidates to solve this problem. All possible series of observations are presented as an oriented graph. Graph vertices are the objects to be observed and graph edges are the time intervals between two successive observations. The Dijksta’s algorithm is used to find the shortest path in the graph. Adding weights to graphic edges allows to optimize other parameters such as the stellar magnitude.

Aerosol model plays an important role in retrieving aerosol properties from satellite measurements. Selecting an appropriate aerosol model usually requires external a priori data like geographic and seasonal aerosol characteristics. Bayesian theory provides a statistical method to evaluate the model performance of a given set of candidate models without the need of external a priori data. We have developed a novel aerosol retrieval algorithm employing a radiative transfer model and the Tikhonov regularization method to estimate aerosol layer height (ALH) and aerosol optical depth (AOD) from the TROPOMI/S5P O2A band (758-771 nm) measurements. In this research, we optimize the algorithm with a Bayesian-based aerosol model selection strategy and apply it to the TROPOMI measurements. The results show that in case of insufficient information for an appropriate micro-physical model selection, the optimized algorithm helps to find the most plausible aerosol model and to improve the accuracy of solutions.

The emission of sea-salt aerosols is governed by various physical processes, with a strong dependence on wind. Strong winds result in the bursting of air bubbles that have been entrained in the waves during whitecap formation and in the tearing of droplets from the wave tops. Marine aerosol particles are considered the dominant contributor to cloud condensation nuclei (CCN) in absence of other aerosol particles over remote oceanic regions. They may also serve as nuclei for the growth of sulphate particles from the oxidation of dimethyl sulphide (DMS). In atmospheric models, the simulated size-resolved emission flux of sea-salt particles is mainly wind-driven. Thus, the assimilation of accurate wind fields is of paramount importance to numerical model performance. In this framework, the launch of the Aeolus satellite by the European Space Agency (ESA) on the 22nd of August 2018, has allowed for the first time the acquisition of three-dimensional wind fields, consisting a significant step-change in the provision of accurate wind profiles on a near-global scale. These observational capabilities are realized through ALADIN (Atmospheric LAser Doppler INstrument), the first Doppler Wind Lidar (DWL) in space, probing the lowermost 30km of the atmosphere to provide profiles of wind, aerosols and clouds along the satellite’s orbital path. The present work aims to investigate possible improvements in sea salt emissions when Aeolus wind profiles are assimilated in regional scale atmospheric models. Towards this objective, two different Weather Research and Forecasting (WRF) model configuration experiments were conducted, each one was initialized with different ECMWF IFS outputs - one with and one without assimilation of Aeolus Rayleigh and Mie L2B wind fields. The region of interest (RoI), extending between 30°W and 60°E and between 5°N and 45°N, encompasses the marine regions of NE Atlantic Ocean and the broader Mediterranean Sea. Our experiments were performed for the period of May 2020 when the intensive EARLINET COVID-19 lidar measurements campaign took place. The sea salt fields from the two experiments are thoroughly evaluated against a variety of independent datasets, including: (a) advanced ground-based remote sensing measurements (i.e. AERONET-CIMEL sun-photometers, NOA’s PollyXT lidar system) acquired at the PANGEA Observatory (Antikythera, SW Greece) during May 2020 and (b) the CALIPSO-based pure-marine product, developed in the framework of the ESA-LIVAS activity. A thorough evaluation of the sea-salt simulated fields, produced with and without the consideration of Aeolus wind assimilation, is in progress. The overarching goals are to justify and interpret the potential improvements on WRF predictive skills, on the representation of sea salt loads, attributed to the assimilation of Aeolus wind profiles. Acknowledgement: "The authors thank the European Space Agency (ESA) and the Hellenic Space Center (HSC) for their support in the framework of the PROgramme de Développement d’Expériences scientifiques (PRODEX) Programme “COllocated wind and aerosol pRofiles of AeoLus for the investigation of ocean sea-salt emissions - CORAL” (PEA 4000131474 ESA-PRODEX-CORAL).

S5p/TROPOMI instrument provides measurements in wide spectral range (from UV to SWIR) and global coverage every day. As it was demonstrated in the frame of ESA S5p+I AOD/BRDF project, these measurements allow retrieval of such aerosol characteristics as aerosol optical depth (AOD),  single scattering albedo (SSA), aerosol size etc. Nevertheless, due to the coarse pixel resolution (7 km by 3.5 km or lower), sensitivity of S5p/TROPOMI measurements to fine spatial variability of aerosol may be strongly limited. This may limit the possibilities of air quality monitoring over urban areas or near aerosol sources. In the frame of ESA CCN “PRISMA+S5p demonstrations for COVID-19 impact” GRASP algorithm was adapted to the combination of S5p/TROPOMI coarse and PRISMA fine spatial resolution measurements. PRISMA (PRecursore IperSpettrale della Missione Applicativa) is an Italian Space Agency (ASI) satellite mission with a swath width of 30 km and a ground sampling distance of 30 m for the hyperspectral sensors (240 channels, with average spectral resolution less than 10 nm). In this presentation we describe the methodology for the combined PRISMA+S5p retrieval with GRASP algorithm and show that this combination allows transferring S5p/TROPOMI advantages of extended aerosol characterization to the high spatial resolution of PRISMA measurements. The results demonstrate a big potential of the approach based on the combination of instruments with coarse and high spatial resolution for aerosol sources identifications and aerosol air quality monitoring at finer scales.

The aerosol optical depth (AOD) is a key parameter for aerosol research owing to its simplicity and suitability for aerosol column amount characterization. In this work, we lay out a satisfactory first attempt to retrieve the AOD at 550nm over land, using a deep neural network which learns from observations collected by the Medium Resolution Imaging Spectrometer (MERIS), on board the Envisat mission (2002-2012), operating on the visible and near-infrared spectral range. The approach has three independent parts, i.e. the preprocessor, the associator and the trainer, whereby the data are significantly reduced in size, matched spatio-temporally with data from the Aerosol Robotic Network (AERONET) and are finally trained with a neural network. The latter makes use of several features involving multispectral reflectances, sun-sensor geometry, and water vapor, all of which are readily available from the Level 1b MERIS product, and no other external information. The retrieved model achieves an overall Pearson correlation coefficient of 0.7 and an absolute error of 0.09 in the validation set for 0.01

The Aerosol Index (AER_AI) as measured by the Tropospheric Monitoring Instrument (TROPOMI) on board the Sentinel-5 Precursor (S5P) platform maintains heritage with aerosol index data records from other satellite instruments including TOMS (EP and Nimbus-7), GOME(-2), OMI and OMPS. The operational TROPOMI AER_AI dataset is available from May 2018 through the present. It is useful not only for tracking ultraviolet (UV) absorbing aerosol plumes of desert dust, volcanic ash, and smoke from biomass burning but also for monitoring the quality of the TROPOMI Level 1b (L1b) data since the AER_AI calculation is very sensitive to the absolute calibration of irradiance and radiance. WIth a nadir pixel size of 3.5 x 5.5 km, TROPOMI’s high spatial resolution also presents specific challenges as non-Lambertian cloud features, cloud shadows, and 3-D effects of clouds are now visible in the TROPOMI AER_AI data. An overview of plans to address these features in future AER_AI updates will be given. This work will also include a description of the impacts on AER_AI due to the observed, wavelength-dependent degradation in the TROPOMI measured irradiance and radiance. Results of recent updates in the L1b data which incorporate corrections for these degradation-driven features will be given.

The Joint Aeolus Tropical Atlantic Campaign (JATAC) is an experimental campaign for the Calibration and Validation (Cal/Val) of Aeolus, an Earth Explorer satellite mission of the European Space Agency. The campaign comprises airborne and ground-based instrumentation in order to provide reference values of the wind profiles, and the aerosols and clouds optical properties for the validation of the Aeolus products. ASKOS (askos.space.noa.gr) is tThe ground-based component of JATAC(named ASKOS) which and was held from July to September 2021 in Cape Verde. ASKOS consists collected of active and passive remote sensing instrumentation measurements for the validation of the Aeolus L2A products of aerosols and clouds, which are (the particle backscatter coefficient, the particle extinction coefficient, and the lidar ratio (irextinction to backscatter ratio) (lidar ratio). One of the deployed instruments is eVe lidar, the ESA’s ground reference system. eVe lidar is a combined linear/circular polarization lidar system with Raman capabilities that operates at 355 nm and retrieves the particle backscatter coefficient, the particle extinction coefficient, and the linear and circular depolarization ratios. eVe is specifically designed to provide the Aeolus mission with ground reference measurements of the aerosols and clouds optical properties. The lidar is implemented in a dual-laser/dual-telescope configuration that allows eVe to simultaneously reproduce the operation of the deployed lidar onboard Aeolus (ALADIN) operating with circularly polarized emission as well as the operation of a traditional lidar system with linearly polarized emission. Targeted measurements of eVe lidar were performed every Friday evening when the closest Aeolus’ orbit overpasses the site of the campaign operations. For the collocated measurements with Aeolus, the eVe lidar was implementing the same pointing geometry of Aeolus in order to reproduce the Aeolus measurements from ground. In total, eight collocated measurements between eVe and Aeolus were collected and the first results from the validation of the Aeolus L2A products will be shown in this work.

Earth’s climate is a complex perturbed system, in which a wealth of chemical, physical and biological processes takes place on a wide range of spatial and temporal scales. A particularly important group of atmospheric processes is termed aerosol-cloud interactions (ACI), which describe cloud adjustments to natural and anthropogenic aerosol particles. Changes in optical and physical properties of clouds are a key factor in both reducing the uncertainty of radiative forcing estimates and in understanding the water cycle. In this work we present retrieval results of the geometrical extent of clouds based on near-infrared oxygen absorption from SCIAMACHY measurements. The retrieved bottom and top altitude of homogeneous clouds are sided by data of fine mode aerosol load and microphysical cloud properties generated within the ESA Climate Change Initiative projects. Global and regional analysis of this parameter’s suite enables the identification of specific ACI regimes. Moreover, we present a technique to inherently account for aerosol perturbation of in-cloud extinction profiles based on the synergistic exploitation of oxygen absorption and multi-wavelength continuum in the solar spectral range. Preliminary results from SCIAMACHY and AATSR show that a more realistic description of in-cloud extinction is beneficial for the accuracy of the geometrical extent of homogeneous clouds. From a future perspective, this retrieval approach can be deployed with measurements of the upcoming NASA Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) mission, scheduled to launch in late 2023. The PACE payload suite comprises the Ocean Color Instrument (OCI), a hyperspectral scanning radiometer, and the polarimetric and multi-angular sensors HARP-2 and SPEXone. The design and the technical complementarity of the three instrument payloads make PACE measurements particularly well-suited for the advancement of our ACI knowledge, whose scientific level of confidence is still quantified as low by the Intergovernmental Panel on Climate Change.

Clouds play a significant role in the Earth’s energy balance and hydrologic cycle through their effects on radiation and precipitation, and therefore are crucial for life on Earth. ENTICE is a proposed satellite with multiple frequency sub-mm microwave radiometers and a cloud radar instrument, which would provide satellite-based global observations of clouds, coincident vertical profiles of cloud ice particle size, ice water content, and in-cloud humidity and temperature. It would help identify the important processes by which clouds evolve and impact climate change, as well as advance our fundamental understanding of cloud processes, leading to reduced uncertainties in climate change predictions. Here, we simulate ENTICE measurements with five different scanning methods: nadir, forward pointing, side scanning, and conical scanning for the radiometers, and nadir pointing for the radar. The satellite’s latitude and longitude were calculated using a two-body force model orbiting the Earth. ENTICE is simulated as flying above a model atmosphere, recording the number and type of clouds seen by the satellite. Our results show that the conical and side scanning methods return the largest number of high-altitude clouds. In conclusion, we find that the satellite would gather enough high cloud samples over its two-year mission to fulfill its science goals, as well as be able to diurnally resolve its data.

The main sources of radiation are solar radiation as incoming short wave (SW) radiation entering the Earth and the longwave radiation (LW) emitted by the Earth's system into space. Absorption of the Sun’s radiation heats the Earth’s surface, which then warms the air above it. Cloud cover is one of the most important atmospheric factors affecting solar radiation, which plays a major role in the energy budget from cooling and heating. Data are taken by cloud cover types and Sunshine Duration in (Hours), Top Net Solar Radiation for Clear Sky in (W/m2), Top Solar Radiation in (W/m2), and Top Incident Solar Radiation in (W/m2) from satellites recorded by the European Centre for Medium-Range Weather Forecasts (ECMWF). The choice of the period (1979-1983) over Baghdad, Mosul, and Basra stations. Otherwise, we have studied analysis monthly mean of LCC, MCC, HCC, and solar radiation types, as well as the relationship between LCC, MCC, HCC, and solar radiation types. The results showed that the relationship between TCC and Top is inverse for the three stations. Where Basra and Mosul stations represent the highest correlation while Baghdad station and represent the middle correlation. As well as the highest amount of TCC occurred in the winter and spring season. Also the highest cloud cover at the Mosul and the lowest cloud cover at Basra.

The Far-infrared Outgoing Radiation Understanding and Monitoring (FORUM) mission, selected for the ESA’s 9th Earth Explorer (EE9) mission, will be launched in 2026. For the first time, spectrally resolved radiance observations covering the Far InfraRed (FIR) band from 100 to 667 cm ⁻¹ with global coverage and with high spectral resolution and radiometric accuracy will be available. This spectral region is highly sensitive to upper-tropospheric–lower-stratospheric (UTLS) water vapor concentration. One of the tools used to support the FORUM proposal during the EE9 competition, and previously, for the data analysis of the REFIR-PAD (Radiation Explorer in the Far InfraRed - Prototype for Applications and Development) balloon observations, is KLIMA (Kyoto protocoL Informed Management of the Adaptation) retrieval code. KLIMA analysed the synthetic FORUM observations in clear sky providing simultaneously the vertical profile of the minor atmospheric gases (such as the water vapour), atmospheric temperature profile, surface temperature, and surface emissivity. During the EE9 phase A selection, KLIMA was employed in various ESA studies such as: FORUM-req for the assessment of the FORUM mission requirements definition; FORUM-E2E for the validation of the End-to-End Simulator (E2ES); FIRMOS for the analysis of the ground-based observations of the FORUM prototype FIRMOS (Far-Infrared Radiation Mobile Observation System). Recently, the KLIMA code was used in the context of the ASI FORUM-science project for the generation of gases optical thicknesses used by the fast code σ-FORUM. In this framework, KLIMA has been upgraded to retrieve also the concentration of the isotopologues of the atmospheric gases. In this work, we will show the assessment of the performance of the retrieval of water vapour isotopologues using the spectral coverage and radiometric accuracy of the FORUM observations, and the comparison with the results obtained when only the middle-infrared information is used. Results both on single measurement and averaged measurements will be shown to evaluate the effect of the radiometric accuracy of the FORUM observations.

The Advanced Infra-Red WAter Vapour Estimator (AIRWAVE) algorithm has been designed to obtain the Total Column of Water Vapor (TCWV) from the measurements of the Along Track Scanning Radiometer (ATSR) instrument series (Casadio et al. 2016, Castelli et al. 2019). On board the Copernicus Sentinel 3 satellite, the Sea and Land Surface Temperature Radiometer (SLSTR) is an advanced version of the ATSR instruments. In the frame of the EUMETSAT project AIRWAVE-SLSTR (https://www.eumetsat.int/AIRWAVE-SLSTR), the AIRWAVE algorithm has been extended to the measurements of SLSTR (AIRWAVE-SLSTR) to produce the SLSTR Level-2 Integrated Water Vapour product. Through this extension, AIRWAVE has the unique capability to retrieve the TCWV over water surfaces for cloud-free scenarios using the SLSTR Thermal Infrared channels, and therefore in both day- and night-time, covering all the water surfaces of the Earth. Moreover, when AIRWAVE-SLSTR will be applied to the whole Sentinel 3 mission, the produced dataset will enable to have a long time series of TCWV obtained with the same algorithm from both (A)ATSR instrument series and SLSTR measurements (therefore from 1991 to 2012 and from 2016 to date). Here we will discuss both the new implementation and the performances of the AIRWAVE algorithm applied to SLSTR measurements using the results of the validation performed with independent products derived from space borne sensors (SSMI/S and MWR) and in situ measurements (IGRA database).

The isotopic composition of water vapour is a useful tracer of moisture transport and moist processes in the atmosphere. On the synoptic time scale, the isotopic composition of atmospheric water vapour is strongly shaped by weather systems and, thus, gives insight into the cycling and distribution of water within such systems. The abundancy of heavy hydrogen (denoted as δD) in atmospheric water vapour can be studied with various tools such as in situ measurements, remote sensing and modelling. In contrast to in situ measurements, satellite retrieved products of δD in water vapour allow to study δD on large spatial scales of several 1000km. The newly developed retrieval of water isotopologues for S5P based on the University of Leicester Full Physics retrieval algorithm provides a dataset to study the distribution and changes of δD on high spatial and temporal resolution. In this study, the S5P δD retrievals are compared to in situ vertical profiles of δD from the L-WAIVE campaign in June 2019 and model simulations with the isotope-enabled weather prediction model COSMOiso. Two large-scale flow situations are assessed for the added value of satellite retrieved total column δD. The combination of in situ measurements with satellite retrieved δD and COSMOiso simulations can help to better constrain vertical δD gradients and the spatial representativeness of in situ δD measurements. Furthermore, recommendations for satellite-model-in situ measurement comparisons of δD are formulated.

Clouds and water vapor are among the most difficult quantities for global climate models to simulate because they are affected by physical processes that operate over scales unresolved by current climate models. We use NASA satellite data to assess the representation of clouds and water vapor structures in 28 climate models that participate in the Coupled Model Intercomparison Project Phase 6 (CMIP6). Each model is assigned numerical scores based on its performance in simulating spatial mean, variance and pattern correlation of multi-year mean clouds and water vapor structures in lower, middle, upper troposphere, and near the tropopause over tropical oceans. We find measurable improvements in CMIP6 models relative to CMIP5 models for both clouds and water vapor. The differences between models and satellite observations and the spread across the models are reduced. In addition, we find that the models’ equilibrium climate sensitivity (ECS) is correlated with overall performance scores for both CMIP5 and CMIP6 models, with a weaker correlation in CMIP6, suggesting that the models that capture better tropical clouds and water vapor distributions tend to have higher ECS. The physical processes responsible for the apparent correlation between ECS and model performance score warrant further study.

The objective of the EC H2020 Copernicus Cal/Val Solution project (CCVS, http://ccvs.eu) is to define a holistic solution for all Copernicus Sentinel missions to overcome current limitations of calibration and validation (Cal/Val) activities. A roadmap will be elaborated to document how the Cal/Val Solution can be implemented, in concertation with European space agencies, Copernicus Services, measurement networks, and international partners. The first step in this process is to identify applicable mission and user requirements and translate them into Cal/Val requirements. This contribution presents the Cal/Val requirements identified for the constellation of Atmospheric Composition (AC) Sentinels being assembled: Sentinel-5p TROPOMI (launched in 2017), and the upcoming series of Sentinel-4 UVN, Sentinel-5 UVNS, and CO2M. A review of generic Cal/Val requirements and methods applicable virtually to every atmospheric composition mission is reported, including the QA4EO framework, generic validation protocols, and maturity assessment tools for validation and data uncertainty. Cal/Val needs specific to the constellation of AC Sentinels are then highlighted in the domains of inter-calibration of sensors, vicarious calibration on natural targets, ground-based validation and Fiducial Reference Measurements, cross-validation of satellite data, field and aerial measurement campaigns, and modelling support. Emphasis is given to new Cal/Val challenges raised by the unprecedented spatial resolution of the Sentinels, the hourly sampling of diurnal changes in atmospheric composition, and operational needs of the Copernicus services for atmospheric composition monitoring (CAMS) and climate change (C3S). The AC Sentinels also contribute to international satellite constellations from which the Copernicus services and applications benefit. Therefore, additional Cal/Val requirements for the AC Sentinels are formulated in view of their interoperability with the IASI-NG and 3MI instruments co-hosted with Sentinel-5 on the MetOp-SG-A platforms, and with other sounders of the CEOS Atmospheric Composition Virtual Constellations for air quality, greenhouse gases, and ozone. The complete documentation is available on the CCVS project website. Acknowledgements This work is supported by the EC H2020 Copernicus Cal/Val Solution project (CCVS) through grant agreement 101004242. References Lambert, J.-C., T. Verhoelst, S. Compernolle, D. Hubert, A. Keppens, B. Langerock, M. K. Sha, F. Tack, K. F. Boersma, G. Tilstra, G. Janssens-Maenhout, and C. Pierangelo, CCVS Atmospheric Composition Missions Cal/Val Requirements, EC H2020 GA101004242, CCVS.BIRA.D1.4, V1.0, 1 June 2021. Alhammoud, B., J.-C. Lambert, G. Hajduch, S. Labroue, A. Meygret, M. Neveu Van Malle, S. Clerc, L. Bourg, M. Hallows, M-L. Denneulin, C. Pierangelo, C. Tison, G. Tilstra, K.F. Boersma, and Y. Govaerts, CCVS Vicarious methods on natural targets, EC H2020 GA101004242, CCVS.ARG.D2.2, V2.1, 2 June 2021. Clerc, S., B. Alhammoud, J.-C. Lambert, G. Hajduch, S. Labroue, C. Pierangelo, C. Tison, F. Bignalet Cazalet, A. Maygret, N. Gobron, Y. Govaerts, M. Raynal, A. Guérou, S. Compernolle, D. Hubert, A. Keppens, and T. Verhoelst, CCVS Inter satellite comparisons, EC H2020 GA101004242, CCVS.CLS.D2.3, V2.0, 31 May 2021. 2.14.0.0

This poster will present the objectives and the first results of the H2020 project Copernicus Cal/Val Solution. The project aims at proposing innovative approaches to improve the performance and optimize Cal/Val activities for the Sentinel Satellites. In a first step, the project has compiled the list of cal/val requirements for Sentinel mission, including atmospheric composition missions such as S5P/TROPOMI and the future missions S4, S5 and CO2M. The project also surveyed reference data used for calibration and validation, from on-board calibration devices to comparisons with models and other satellites, vicarious methods using natural scenes, as well as systematic and campaign-based in-situ measurements. In a second phase recently started, we identify current gaps, and propose innovations and rationalization to improve cal/val such as the definition of mutli-disciplinary "supersites".

Calibration/Validation activities are crucial to the success of the increasing number of Earth observation satellite missions. Different ground-based networks (e.g. in Europe AERONET, EARLINET, PANDONIA, etc.) are collecting long-term records of atmospheric information for, among other goals, supporting the needs of those missions. Also, initiatives like ACTRIS (Aerosols, Clouds, and Trace gases Research InfraStructure) devoted to join efforts and homogenize the networks and their protocols are created. Here we present the DIVA project (prototype) that is a hub to collect, handle, archive, and exploit in a synergetic way observational data from ground and space that aims at validating ESA and Copernicus satellite missions and scientific analysis. The system is versatile and integrates ground-based (lidar, photometer, and spectrometer), satellite and transport model data. It can run stand-alone and synergetic algorithms for advanced data products, use the data from different platforms and sensors, as well as include innovative data mining and data visualization tools. DIVA proposes an easy, useful and practical python interface to manipulate the data. The system runs in the cloud and it is especially convenient to handle satellite data because it avoids the necessity of downloading large datasets. GRASP (Generalized Retrieval of Atmosphere and Surface Properties) has been selected as the main algorithm to exploit synergies of the data included in DIVA and retrieve advanced properties from the combination of several instruments. GRASP is currently employed for lidar, photometer, and spectrometer data providing a combined retrieval of aerosol and gas properties. The first prototype of DIVA has been already implemented and it uses the selected algorithms to treat the currently available data. It is expected that the system grows up and it is offered to the users for creating a community that benefits from it.

The Fundamental Data Record for ATMOSpheric Composition (FDR4ATMOS) project is part of the ESA Long Term Data Preservation (LTDP) programme and has two objectives: First, update the SCIAMACHY processing chain for better Ozone total column data: After the full re-processing of the SCIAMACHY mission with the updated processor versions, the validation showed that the total Ozone column drifted downward by nearly 2% over the mission lifetime. This drift is likely caused by changes in the degradation correction in the Level 1 processor, that led to subtle changes in the spectral structures. These are misinterpreted as an atmospheric signature. FDR4ATMOS updated the Level 0-1 processor accordingly and a full mission re-processing hast started. In an addition to the original plan, we also incorporated lunar data in the SCIAMACHYLevel 1b product. The instrument performed regular lunar observations buiding up a unique 10 year data set of lunar spectra from the UV to the SWIR with moderately high spectral resolution. In the project we calibrated these data. The results show an agreement within the error with other lunar data (e.g. ROLO). The second objective of the FDR4ATMOS project is to develop a cross-instrument Level 1 product for GOME-1 and SCIAMACHY for the UV, VIS and NIR spectral range, with focus on the spectral windows used for O3, SO2, NO2 total column retrieval and the determination of cloud properties. FDR4ATMOS aims to build a Fundamental Data Record (FDR) of Level 1 products, i.e. radiances and reflectances. The GOME-1 and SCIAMACHY instruments together span 17 years of spectrally highly resolved data essential for air quality, climate, ozone trend and UV radiation applications. We plan to generate harmonised data sets that allows to directly use it in long-term trend analysis, independently of the instrument. Since this was never done for highly resolved spectrometers, new methods have to be developed that e.g. take into account the different observation geometries and observation times of the instrument without impacting the spectral structures that are used for the retrieval of the atmospheric species. The resulting algorithms and the processor should also be as generic as possible to be able to easily transfer the methodology to other instruments (e.g. GOME-2 and Sentinel-5p) for a future extension of the time series. We will present the current status of the project, including results for the updated SCIAMACHY processor, the uncertainty analyses and will report the status of the data analysis for the harmonisation of data.

Now in operational production for almost three years, Sentinel-5p TROPOMI NO2 measurements are proving invaluable for a plethora of applications on air quality, not in the least to quantify the impact of increasingly stringent air quality policies, or of the lockdown measures put in place in 2020 to halt the spread of the SARS-CoV-2 pandemic. At the same time, several improvements have been implemented in the operational production of the NO2 data. Major updates were the improved determination of the cloud top pressure in the auxiliary FRESCO(-S) cloud retrieval (since NO2 processor v1.4, as of November 29, 2020), impacting the derived tropospheric NO2 column amounts, and the use of a new version of the underlying L1b data since processor v2.2 (as of July 1, 2021). Further developments have also been undertaken by the community, such as the use of profile information extracted from the CAMS-regional ensemble to replace that of the TM5 assimilation system. The impact of these improvements on the data quality deserves a thorough quantification as any change –for instance in bias- may skew derived trends. The quality of the official S5P-TROPOMI data offered to the public is assessed within the S5P MPC, in particular by ground-based validation in the Validation Data Analysis System (VDAF), complemented with input from the AO project NIDFORVAL - and augmented with analyses of diagnostic data sets and scientific products. Essential in these assessments is the comparison to Multi-Axis (MAX-) and Zenith-sky scattered light (ZSL-) DOAS measurement from the Network for the Detection of Atmospheric Composition Change (NDACC), and to Pandonia Global Network (PGN) measurements. While the overall quality of the operational S5P NO2 data is well established (and documented), we report here specifically on (1) the evolution of the data quality through the various operational processor upgrades, and (2) the improvement in consistency between satellite and ground-based measurements when replacing the TM5 vertical NO2 profiles with those from the CAMS-regional ensemble. In short, we find at many sites a reduction in the amplitude of the negative mean difference for the tropospheric columns , from about -30% to -20%, by the introduction of the updated cloud top pressures in v1.4, and from about -7% to -4% for the stratospheric columns by the introduction of the new L1b data (v2.2). These changes imply a discontinuity in the bias properties and data before and after should only be combined with great caution, at least until a full reprocessing of the NO2 data set with the upgraded processor is carried out.         The use of CAMS-regional vertical profiles improves the agreement with the ground-based data in terms of bias, but - contrary to expectations - it does not improve the correlation, nor does it reduce the dispersion of the differences. Further work is thus needed to elucidate fully the origin of the differences between ground-based and TROPOMI NO2 data.

The Sentinel-5 Precursor (S5P) mission represents the first in a series of atmospheric sounding systems within the Copernicus programme. The S5P mission is a single-payload satellite in a low Earth orbit that provides daily global information on concentrations of trace gases and aerosols important for air quality, climate forcing, and the ozone layer.The payload of the mission is the TROPOspheric Monitoring Instrument TROPOMI, which has been jointly developed by the Netherlands and ESA. The instrument contains four spectrometers, divided over two modules sharing a common telescope, measuring the ultraviolet, visible, near-infrared and shortwave infrared reflectance of the Earth. The imaging system enables daily global coverage using a push-broom configuration, with a spatial resolution as low as 5.5 x 3.5 km2 in nadir from a Sun-synchronous orbit at 824 km and an equator crossing time of 13:30 local solar time. Once per day the solar spectral irradiance is measured via a dedicated calibration port in the instrument.After more than 4 years in orbit, a combination of instrument ageing and drifts makes it necessary to update radiometric calibration key data also during the nominal operations phase. We will report on the in-flight calibration approach to characterize and differentiate between the various degradation and drift effects. The analysis of the in-flight data results in calibration key data which is used to correct these effects in the Level 0-1b processor. Depending on the source of the degradation within the instrument, a correction is applied to the irradiance data or also the radiance data. The observed optical degradation is most pronounced in the short wavelength range of the instrument.We will report on the current status of the known ageing and drift effects in TROPOMI and how well they can be addressed by corrections in Level 0-1b processing.

TROPOspheric Monitoring Instrument (TROPOMI) was launched in October 2017 on board of the ESA/Copernicus Sentinel-5 Precursor (Sentinel-5p) satellite. This UV/VIS/NIR/SWIR nadir sounder measures daily, on the global scale and at unprecedented horizontal resolution, the Earth radiance and the atmospheric abundance of species related to air quality, climate forcing, stratospheric ozone, UV radiation and volcanic hazards. At the vanguard of the upcoming series of Sentinel-4 and Sentinel-5 atmospheric composition missions, Sentinel-5p TROPOMI contributes data to Copernicus information services on our environment and security - e.g., through data assimilation in the Copernicus Atmosphere Monitoring Service (CAMS) operated by ECMWF. Since the beginning of its routine operations, the quality of the TROPOMI data products has been evaluated and monitored by a specific validation team cooperating in the framework of the Sentinel-5p Mission Performance Centre. Documented on a dedicated portal (https://mpc-vdaf.tropomi.eu/), the routine operations validation service supplied by this team relies largely on the Validation Data Analysis Facility (VDAF) and its Automated Validation Server (VDAF-AVS). This contribution browses the status of the architecture, data streams and functionalities of the VDAF-AVS with a focus on its application to Sentinel-5p. The VDAF-AVS evaluates TROPOMI data routinely with respect to Fiducial Reference Measurements (FRM) and other validation data archived at the ESA Atmospheric Validation Data Centre (EVDC). FRM and other validation data are collected through several projects of the ESA FRM programme, and from ground-based monitoring networks archives: Cloudnet, NDACC, PGN, SHADOZ, TOLNET, and WOUDC. TROPOMI and validation data are ingested and inter-compared in an automated validation system grounded on generic validation protocols and metrics, and implemented using the Atmospheric Toolbox (CODA and HARP toolsets) developed in heritage projects and tailored to the TROPOMI validation needs and the Copernicus Sentinels operational requirements. The public output of the system is available on https://mpc-vdaf-server.tropomi.eu/ Currently, the VDAF-AVS provides routinely validation results for the following TROPOMI operational data products: O₃ (total and tropospheric columns), NO₂ (total, tropospheric and stratospheric columns), total columns of HCHO, CO and CH₄, and cloud (top) height. The VDAF-AVS is also ready for routine validation of O₃ profile data. Further integration of other data products (e.g., SO₂ column and aerosols) and of other validation data sources (e.g., EUBREWNET, TCCON and COCCON) is in progress. The past 5 years of VDAF-AVS implementation, operation and evolution constitute a pathfinder for the operational validation of the upcoming constellation of atmospheric composition Sentinels. 2.14.0.0

We describe the new set-up of the high resolution regional atmospheric chemistry model WRF-Chem over western Europe and evaluate the NO2 VCDs against TROPOMI observations. The model's horizontal resolution (3x3 km2) is comparable to that of TROPOMI in the nadir view. We observe very good initial model performance if monthly mean spatial patterns of the simulated NO2 VCDs are compared against TROPOMI observations (Pearson correlation coefficient = 0.9). Reprocessing of the TROPOMI offline product to replace the coarse resolution a priori from TM5 by WRF-Chem increases the VCDs by 13% over the complete domain but up to 30% over the hotspots. However, further adjustments (discussed below) had to be made in order to improve the agreement with daily measurements. Previous studies using the WRF-Chem model sampled the model output at the fixed UTC hour close to the equator overpass for comparison with satellite observations. We show that the actual overpass at a given location can occur within a time span of up to 5 hours on different days for different TROPOMI orbits. This can lead to large artefacts due to sampling errors for short-lived species such as NO2. Hence, we perform the comparison for individual orbits with the model output of the nearest UTC hour at a given pixel to limit the difference by a maximum of 30 min. As expected, improved agreement for individual orbits was evident, and individual plume evolution could be tracked when there were two overpasses over some regions of the domain in a single day. We also demonstrate the improvement in daily comparisons if the model is nudged towards the ERA5 reanalysis dataset of temperature and zonal and meridional components of winds. We use spectral nudging, such that only the large-scale features (> 1000 km) are relaxed towards the reference to avoid large drifts. The application of spectral nudging improved the agreement for individual plume directions.

Atmospheric moisture is a key factor for the redistribution of heat in the atmosphere, and there is a strong coupling between atmospheric circulation and moisture pathways which is responsible for most climate feedback mechanisms. Water isotopologues can make a unique contribution for better understanding this coupling. In recent years, water vapour isotopologue observations from satellites have become available from thermal nadir infrared measurements (TES, AIRS, IASI) which are sensitive above the boundary layer and from shortwave-infrared (SWIR) sensors (GOSAT, SCIAMACHY) that provide column averaged concentrations including sensitivity to the boundary layer. Sentinel 5P (S5P) measures SWIR radiance spectra that allow retrieval of water isotopologue columns but with much improved spatial and temporal coverage compared to other SWIR sensors providing a large potential for scientific and operational applications. We have developed a new retrieval of water isotopologues for S5P, specifically of the ratio of HDO/H2O, based on the University of Leicester Full Physics (UoL-FP) retrieval algorithm and examined and characterized the retrieval performance by the validation of retrieved water isotopologues against reference data sets (MUSICA NDACC data and TCCON) and by intercomparisons to satellite data from IASI and GOSAT. The impact of the satellite derived water isotopologues data is assessed against in situ measurements of vertical water isotopologues profiles and model calculations from the isotope-enabled numerical models ICON-ART-ISO and COSMOiso . In this presentation, we will first describe the retrieval approach and the derived S5P data products results of the validation and satellite intercomparison, and we present model-satellite case studies for Central Europe and West Africa. We will conclude with an outlook outlining future development steps

An extension of the scientific HDO/H2O column data product from the Tropospheric Monitoring Instrument (TROPOMI) including clear-sky and cloudy scenes is presented. The retrieval employs a forward model which accounts for scattering, and the algorithm infers the trace gas column information, surface properties and effective cloud parameters from the observations. The extension to cloudy scenes greatly enhances coverage, particularly enabling data over oceans. The data set is validated against co-located ground-based Fourier transform infrared (FTIR) observations by the Total Carbon Column Observing Network (TCCON). The median bias for clear-sky scenes is 1.4 × 1021 molec cm−2 (2.9 %) in H2O columns and 1.1 × 1017 molec cm−2 (−0.3 %) in HDO columns, which corresponds to −17 ‰ (9.9 %) in a posteriori δD. The bias for cloudy scenes is 4.9 × 1021 molec cm−2 (11 %) in H2O, 1.1 × 1017 molec cm−2 (7.9 %) in HDO, and −20 ‰ (9.7 %) in a posteriori δD. At low-altitude stations in low and middle latitudes the bias is small, and has a larger value at high latitude stations. At high altitude stations, an altitude correction is required to compensate for different partial columns seen by the station and the satellite. The bias in a posteriori δD after altitude correction depends on sensitivity due to shielding by clouds, and on realistic prior profile shapes for both isotopologues. Cloudy scenes generally involve low sensitivity below the clouds, and since the information is filled up by the prior, it plays an important role in these cases. Over oceans, aircraft measurements with the Water Isotope System for Precipitation and Entrainment Research (WISPER) instrument from a field campaign in 2018 are used for validation, yielding a bias of −3.9 % in H2O and −3 ‰ in δD over clouds. To demonstrate the added value of the new data set, a short case study of a cold air outbreak over the Atlantic Ocean in January 2020 is presented, showing the daily evolution of the event with single overpass results.

In this paper, we present the total column water vapour (TCWV) retrieval for the TROPOspheric Monitoring Instrument (TROPOMI) observations in the visible blue spectral band. The retrieval was first developed to retrieve TCWV from Global Ozone Monitoring Experience 2 (GOME-2). We have modified the settings of the retrieval to adapt it for TROPOMI observations. The TROPOMI TCWV algorithm follows the typical two steps approach with spectral fit retrieval of slant columns and converts the slant columns to vertical columns using air mass factors (AMFs). An iterative optimization algorithm is developed to dynamically find the optimal a priori water vapour profile for AMF calculation. Further optimization on the spectral retrieval and air mass factor calculations are implemented. Details of the TCWV retrieval are presented.The TCWV retrieval algorithm is applied to TROPOMI observations from May 2018 to May 2021. The results are validated by comparing to ERA5 reanalysis data, GOME-2, MODerate resolution Imaging Spectroradiometer (MODIS) and Special Sensor Microwave Imager Sounder (SSMIS) satellite observations. TCWV derived from TROPOMI observations agree well with the other data sets with Pearson correlation coefficient (R) ranging from 0.96 to 0.99, and with bias ranging from -3.25kg/m2 to 0.36kg/m2. The small discrepancies found between TROPOMI and reference data sets are related to the differences in measurement technique, measurement time, surface albedo issue, as well as cloud and aerosol contamination. This study demonstrated that the algorithm can provide stable and consistent results and can be applied to the current and forthcoming satellite sensors.

Atmospheric water plays a key role for the Earth’s energy budget and temperature distribution via radiative effects (clouds and vapour) and latent heat transport. Thus, the distribution and transport of water vapour are closely linked to atmospheric dynamics on different spatiotemporal scales. In this context, global monitoring of the water vapour distribution is essential for numerical weather prediction, climate modelling, and a better understanding of climate feedbacks. Total column water vapour (TCWV), or integrated water vapour, can be retrieved from satellite spectra in the visible “blue” spectral range (430-450nm) using Differential Optical Absorption Spectroscopy (DOAS). The UV-vis spectral range offers several advantages for monitoring the global water vapour distribution: for instance, it allows for accurate, straightforward retrievals over ocean and land surface even under partly-cloudy conditions. To investigate changes in the TCWV distribution from space, the Ozone Monitoring Instrument (OMI) on board NASA’s Aura satellite is particularly promising as it provides long-term measurements (late 2004-ongoing) with daily global coverage. Here, we present a global analysis of trends of total column water vapour retrieved from multiple years of OMI observations (2005-2020) and put our results in context to TCWV trends from other climate data records (e.g. reanalysis models or satellite measurements). Moreover, we investigate if the assumption of constant relative humidity over climatological time periods is valid and consequently if the changes in TCWV follow a Clausius-Clapeyron response (i.e. if the water vapour concentration changes by approximately 7% per Kelvin). In addition, we also demonstrate that the OMI TCWV data set can also give insights into changes of the global atmospheric circulation.

Water vapour is a crucial component of the Earth climate system. As the most significant natural contributor to the greenhouse effect, it also has the capacity to regulate evaporative and transpiration processes. Therefore, water vapour is closely connected to both the global hydrological cycle and energy budget. The time water spends in the atmosphere between evaporation and precipitation is directly linked to moisture transport, extreme precipitation and hydrological sensitivity due to climate change. This study investigates the residence time of water vapour from satellite, reanalysis and model ensembles to evaluate our current understanding and assess the hydrological response over a 26 year period.

ESA’s Climate Change Initiative (CCI) was established to tackle the challenges encountered in merging climate data records (CDRs) of Essential Climate Variables and to provide climate modelers and researchers with stable long-term records from current and past European (and third-party) satellite missions. In this contribution, we focus on two vertically resolved water vapour data records recently released by the CCI Water Vapour project and we assess their potential for climate applications. The first CCI H2O CDR represents monthly zonal mean profile data in the stratosphere and lower mesosphere since the mid-1980s, combining observations by limb and occultation sounders. The second CCI H2O CDR is constructed from nadir and limb measurements allowing to obtain latitude- and longitude-resolved monthly mean profiles during 2010-2014 covering the troposphere and the UTLS region. We start with an overview of the input profile data, the merging principles and the general characteristics of these CCI data records, and relate these to those of other water vapour CDRs developed by NOAA (SWOOSH) and NASA (GOZCARDS). This provides the contextual information to carry out an assessment of the coherence of vertically resolved H2O CDRs. Multiple linear regression was applied to all time series to infer parameters of seasonal cycle, Quasi-Biennial Oscillation, trends and multi-annual mean water vapour. We present the zonal and vertical structure of these patterns and discuss the (dis)agreement between the CDRs. We conclude that the various periodic patterns are represented in a similar way; the trends, however, can differ considerably, depending on period and region in the atmosphere. These differences are interpreted in terms of, e.g., differences in contributing sensors and data versions. In conclusion, the ESA CCI vertically resolved water vapour data records offer a valuable new source of information for global, long-term studies and climate applications.

The Simultaneous Atmospheric and Cloud Retrieval (SACR) code is an algorithm comprising a forward and a retrieval model capable to simultaneously retrieve the main surface and atmospheric variables and of the cloud optical and microphysical properties, from infrared spectral radiance measurements. These atmospheric parameters are relevant in climate change studies. In particular, the characterization of cirrus cloud radiative properties represents a paramount goal, since this type of clouds modulates the incoming solar radiation and the thermal outgoing emission from the ground. Furthermore, cirrus clouds cover permanently more than 30% of the whole planet surface reaching 70% in the tropical regions. Despite their importance in the Earth Radiation Budget (ERB), cirri are difficult to detect, especially in polar regions, and their properties are still affected by large uncertainties. It has been recently recognized that the spectral region between 100 and 1600 cm-1 (6.2-100 μm), and in particular the far infrared (FIR) region below 600 cm-1, contains key information about the atmospheric water vapour and ice clouds. Despite this, there is a lack of spectrally resolved measurements covering the FIR region, either from air/space borne and from ground-based sensors. To cover this observational gap, the Far infrared Outgoing Radiation for Understanding and Monitoring (FORUM) mission was selected by the European Space Agency (ESA) as the 9th Earth Explorer (EE-9). For the first time from space, FORUM will acquire spectrally resolved measurements of the Earth emission. The procedure to retrieve relevant atmospheric parameters from FIR and MIR simulated spectral radiances (as those that will be measured by FORUM) is discussed in this work. Specifically, the parameters considered in the retrieval are: the water vapour and temperature profiles, the cloud optical depth, effective diameters of the particle distributions, the ice fraction in the presence of mixed phase clouds, the cloud position and the surface temperature. The performance of the retrieval is examined on the basis of a few test cases.

The Earth’s Outgoing Longwave Radiation (OLR) spectrum from 100 to 1600 cm-1 that will be measured by FORUM (Far-infrared Outgoing Radiation Understanding and Monitoring), the ninth Earth Explorer mission of ESA, planned for the launch in 2026, can be exploited to retrieve the longwave spectral fluxes. In this study, we consider the approach that calculates the spectral flux from the atmospheric state retrieved from the spectral radiance measurements. The atmospheric state (mainly, vertical profiles of humidity and temperature, and ozone column), cloud and surface properties are retrieved using an inversion model based on a line-by-line radiative transfer code with the capability of simulating the multiple scattering. The retrieved parameters are then used to simulate the hemispherical spectral emission at the top-of-atmosphere that can be angularly integrated to give the spectral flux, potentially extending the spectral range to include the whole OLR. The radiance is assumed to be homogeneous in azimuth and the integral over the zenith angle is solved using a Gaussian quadrature. If the spectral radiance measurement contains most of the emitted energy, as in the case of FORUM, the error due to the conversion of radiance to flux can be very small. We apply this approach to one of the few existing set of measurements covering the wide-spectral range of FORUM in conditions similar to space observations, i.e. from a stratospheric balloon platform. This measurement was performed in 2005 from Teresina, Brazil, by the Radiation Explorer in the Far InfraRed – Prototype for Applications and Development (REFIR-PAD) instrument that was operated to measure at nadir the spectral radiance from 100 to 1400 cm-1 with a spectral resolution of 0.5 cm-1. The flight lasted approximately 8 hours acquiring measurements from an average floating altitude of 34 km in clear sky conditions. From these measurements, the vertical profiles of humidity and temperature and the surface temperature were retrieved using a few inversion codes based on forward models developed at our own premises (KLIMA, SACR, and σ-FORUM). Spectral fluxes were then computed via a quadrature approach. From spectral fluxes, band integrated fluxes can be finally calculated and compared with the results from the Rapid Radiative Transfer Model (RRTM). We discuss the advantages and the disadvantages of these methods and analyse the limits of their applicability.

In cold and dry conditions typical of high-latitude and high-altitude locations, a significant fraction of the far-infrared (FIR) emission from the surface can reach the top of atmosphere (TOA). Recent works examined the impact of FIR surface emissivity on climate model projections and highlighted the influence of surface spectral emissivity on the surface and TOA FIR radiation budgets of polar regions. As spectral measurements in the FIR are still scarce, the authors of these studies used theoretical modelling to generate representative surface spectral emissivities. FORUM (Far-infrared Outgoing Radiation Understanding and Monitoring) is expected to be launched in 2026, and its core instrument will measure across Earth’s FIR part of the electromagnetic spectrum. One of the key objectives of the mission is to obtain an extensive database of surface emissivity in the FIR. During the mission phases A and B1, the Far-Infrared Radiation Mobile Observation System (FIRMOS), a Fourier transform spectrometer, was built as an instrument demonstrator. In the future FIRMOS may also contribute to FORUM validation from the ground and from stratospheric balloon platforms. In the end of 2018, FIRMOS was deployed for a two-month ground-based campaign on Mount Zugspitze (German Alps, 2962m a.s.l.) where it measured the downwelling atmospheric spectral radiance from 100 to 1000 cm-1 (10-100 µm) with a resolution of 0.3 cm-1. During this campaign, upwelling snow surface emission spectra were also acquired with FIRMOS. A variety of 9 snow and ice samples were collected in the vicinity of the experiment site. The samples were characterised in terms of snow grain type, density (kg m−3), and specific surface area (SSA, m2 kg−1). The radiance measurements of the samples and of the atmosphere were inverted to retrieve spectral emissivity by means of an algorithm based on optimal estimation. This presentation describes the modelling of the inverse problem, the retrieval method and the preliminary results obtained.

In a climate change era, measuring atmospheric parameters is fundamental for monitoring the status and predicting the patterns of the atmosphere. For this reason, several space missions, already launched or to be launched by different space agencies are devoted to these measurements. While most key parameters are or will be available, some others are still missing. Among them, the water vapor (WV) in the bottom part of the troposphere, which can contribute to improve numerical weather prediction models (NWP) on short time scales. In order to fill these gaps, a method for obtaining the integrated water vapor (IWV) along a microwave link and based on a pair of attenuation measurements, called NDSA (Normalized Differential Spectral Attenuation), was proposed years ago. The NDSA method is based on measuring a parameter called "spectral sensitivity", which is the normalized incremental ratio of the spectral attenuation and was found to be linearly related to the IWV along the radio-link path. Some studies, supported by ESA, have shown the NDSA capability to effectively estimate the IWV along the path between two Low Earth Orbit (LEO) satellites - one carrying a transmitter, the other a receiver - in a limb measurement geometry, using transmission frequencies in the Ku and K bands for paths crossing the low troposphere (below 10 km), or much higher in case higher tropospheric layers are involved. The ESA studies focused on the retrieval of vertical profiles of IWV and WV in the case of LEO satellites orbiting in the same plane, but along opposite directions (counter-rotating), and only marginally the very different case of the application of the NDSA technique to satellites orbiting in the same plane and along the same direction (co-rotating). The latter is the concept of a recent project, SATCROSS, supported by the Italian Space Agency: the objective is providing a pre-feasibility study for a space remote sensing system based on a train of co-rotating LEO satellites. In such a configuration, in which one or more LEO satellites with an on-board transmitter follow(s) one or more LEO satellites with receiving apparatuses, the NDSA measurements refer to links that cross the troposphere at certain altitudes and “brush” an entire annular region in the orbital plane. It is thus possible to estimate the two-dimensional field of WV in the aforementioned annular region starting from the entire set of IWV measurements based on an inverse problem formulation. Specific SATCROSS activities are: 1) analysis and application of appropriate two-dimensional inversion algorithms; 2) development of an end-to-end simulator; 3) analysis of the impact of the WV products on NWP; 4) definition of the mission characteristics and of the satellites payload, based on Cubesat technology. As a support to these activities, a measurement campaign is being carried out in a ground-to-ground configuration with an instrument able to provide NDSA measurements at 19 GHz. The purpose of this presentation is to illustrate the overall framework of the SATCROSS project, while the details of the related activities are given by companion presentations in ATMOS 2021.

Atmospheric carbon dioxide (CO2) and methane (CH4) measurements provide valuable information about the sources that emit the critical greenhouse gases (GHGs) into the atmosphere and the sinks that remove them. At global scales, atmospheric concentrations of CO2, CH4 and other well-mixed GHGs are well characterized by precise, ground-based and airborne in situ measurements. Recent advances in space-based remote sensing methods are providing new opportunities to augment the resolution and coverage of those ground and airborne measurements with spatially-resolved estimates of the column-averaged CO2 and CH4 dry air mole fractions, XCO2 and XCH4. The current generation of space-based sensors includes those carried by Japan’s Greenhouse gases Observing SATellite (GOSAT) and GOSAT-2, the National Aeronautics and Space Administration (NASA) Orbiting Carbon Observatory-2 (OCO-2) and OCO-3 missions and the Copernicus Sentinel 5 Precursor missions. These ground-based, airborne and space-based datasets are being analyzed with advanced atmospheric inverse modeling systems to yield top-down estimates CO2 and CH4 fluxes on spatial scales spanning individual power plants or large urban areas to nations, biomes or the entire planet. Carbon cycle scientists are using these results to study a broad range of topics ranging from the response of tropical ecosystems to the intense 2015-2016 El Niño to the impact of the COVID-19 lockdowns on anthropogenic CO2 emissions. Exploiting these developments, the CEOS/CGMS Joint Working Group on Climate GHG Task team is coordinating the development of pilot, country-scale budgets of CO2 and CH4 emissions and removals to support the first global stocktake of the UNFCCC Paris Agreement. These top-down budgets are not as process specific as the bottom-up GHG emission inventories compiled by national inventory agencies, but complement those products by providing an integrated constraint on the net amount of each gas that is exchanged between the surface and the atmosphere by natural and anthropogenic processes. The principal objective of this effort to start conversation with stakeholders in the UNFCCC and policy communities, the national inventory agencies, and the GHG measurement and modeling communities to refine the requirements for a purpose-built global carbon monitoring system that fully exploits future space-based capabilities such as NASA GeoCarb, Japan’s GOSAT-GW, and the Copernicus CO2M missions.

The Fast atmOspheric traCe gAs retrieval (FOCAL) algorithm has been originally developed for OCO-2 retrievals with the focus on the derivation of XCO2. FOCAL is a candidate algorithm for the forthcoming CO2M mission. Recently, the FOCAL method has been also successfully applied to measurements of the Greenhouse gases Observing SATellite (GOSAT) and its successor GOSAT-2. In this presentation we will show new results from updated GOSAT and GOSAT-2 FOCAL retrievals. This will include also results for gases other than XCO2, e.g. XCH4, water vapour or CO.

With atmospheric methane concentrations increasing at record pace, identifying the emission sources with the largest mitigation potential is paramount. The Tropospheric Monitoring Instrument (TROPOMI) aboard ESA’s Sentinel-5p satellite was launched in 2017 and provides daily global coverage of methane concentrations at up to 7 x 5.5 km2 resolution. These data can be used to detect persistent methane emissions as well as large emission events such as accidents in the natural gas industry. We detect emissions coming from these activities using a convolutional neural network trained to detect plume-like features in the TROPOMI data combined with a support vector classifier to filter for artefacts. This automated machine learning approach allows constant global monitoring. While some plumes can be attributed to a specific source using just TROPOMI data, its kilometer-scale resolution usually does not enable that. Therefore, we use super-emitter detections made based on TROPOMI data to guide high-resolution instruments such as GHGSat, Sentinel-2, or PRISMA to find the exact source responsible. For persistent sources, we use long-term TROPOMI data combined with metrological data to pinpoint the area the source is located in as precisely as possible. The meter-scale observations of high-resolution instruments over limited domains can then be used to identify the exact facilities responsible for the enhancements seen in TROPOMI and estimate their emissions. This information can subsequently be used to inform the operators allowing – for example – gas leaks to be fixed. We also use the TROPOMI detections of one-time events to focus data analysis of Sentinel-2 observations which provide global coverage every five days. We will show examples of these synergies applied to persistent emissions from coal mines and landfills and both persistent and transient emissions in the oil and gas industry. This combined use of satellite instruments gives us an important tool to enable and evaluate emission mitigation.

Methane (CH4) is an important greenhouse gas, which accounts for the second-largest share of radiative forcing caused by human activities since pre-industrial times. It has a much shorter atmospheric lifetime and a considerably higher global warming potential than the most important anthropogenically modified greenhouse gas, carbon dioxide (CO2). Hence, a combined climate change mitigation strategy, aiming at reducing both CO2 and CH4 emissions in parallel, addresses long-term and near-term effects of global warming and is required to achieve climate goals most efficiently.The TROPOspheric Monitoring Instrument (TROPOMI) onboard the Sentinel-5 Precursor satellite, which was successfully launched in October 2017, is a spaceborne nadir-viewing imaging spectrometer measuring solar radiation reflected by the Earth in a push-broom configuration. It has a wide swath on the terrestrial surface and covers wavelength bands between the ultraviolet (UV) and the shortwave infrared (SWIR), combining a high spatial resolution with daily global coverage. Abundances of the atmospheric column-averaged dry air mole fractions XCH4 are retrieved from TROPOMI's radiance measurements in the 2.3 µm spectral range of the SWIR part of the solar spectrum using the scientific retrieval algorithm Weighting Function Modified Differential Optical Absorption Spectroscopy (WFM-DOAS).TROPOMI's unique combination of high precision, accuracy, and spatio-temporal coverage enables the systematic detection of sufficiently large emission sources in a single satellite overpass. We present a method for the quantification of hot spot emissions based on daily recurrent TROPOMI observations and apply it to determine local anthropogenic methane emissions including examples from the oil, gas, and coal industry in North America, Central Asia, and Europe.

The Northern high latitude regions are changing: the rapid increase in temperature, resulting in a changing cryosphere and increasing human activity, is potentially increasing high-latitude methane emissions. This remote region is best monitored using satellites but poses several challenges for greenhouse gas satellite observations due to the lack of solar radiation during winter, frequent cloud coverage, and long-standing snow coverage on the ground. The column-averaged methane (XCH4) observations from TROPOMI onboard Sentinel 5P satellite provide unprecedented coverage in this region, compared to previous satellite missions or in situ measurements. With the aim to ultimately assess their added value for methane source estimation, we present here a systematic comparison of three TROPOMI methane products: operational and scientific SRON products, and the scientific WFMD product. We focus exclusively on latitudes poleward of 50°N and evaluate the XCH4 biases at four high-latitude TCCON sites: Ny Ålesund, Eureka, Sodankylä and East Trout Lake. The accuracy and precision of all three products compare well with the TCCON, although a persistent seasonal bias in TROPOMI XCH4 (high values in spring, low values in autumn) is found for all three satellite products. We quantify these biases and study their possible origins by analysing the albedo effects from snow cover and the changes in the CH4 profile shape caused by high-altitude depletion of methane in the polar vortex. Comparisons to atmospheric profile measurements with AirCore carried out in Northern Finland support the analysis and help validate prior profiles used in the retrievals. In addition, we evaluate the seasonal coverage over the continuous and discontinuous permafrost regions, reflecting the potential of TROPOMI to inform inversion models on changing emissions in these regions. We also present a comparison of inverse model results from Carbon Tracker CTE-CH4, which show that these seasonal biases may have a significant impact on the fluxes. We conclude that the availability of several XCH4 products based on different retrievals is important for a more reliable interpretation of spatial patterns, anomalous values and understanding of the potential origin of biases.

The ESA Methane+ project investigates synergies between SWIR and TIR retrieval approaches, using data from TROPOMI and IASI, and their application in global inverse modelling of methane. The consistency of the information provided by SWIR and TIR retrievals is evaluated using vertical profile information from the TM3 and TM5 models, supported further by independent in situ measurements. This approach is important for SWIR and TIR retrievals, because of their different vertical sensitivities, complicating a direct comparison of retrievals. The use of two transport models, and two retrieval datasets for the SWIR (RemoTeC and WFMD) and TIR (LMD and RAL) retrievals, helps in distinguishing robust variations in methane from uncertainties in transport models and satellite retrievals. In addition, inversions using the surface network and the operational TROPOMI retrieval are used as references. Global methane inversions have been performed using the four retrieval datasets for a two-year period starting from the beginning of the operational processing of TROPOMI data in spring 2018, until the first months of 2020. The comparison between inversion results for 2019 and 2020 is interesting in particular, because of the exceptional rapid increase in global CH4 during the COVID-19 pandemic reported by the global surface measurement network. Inversions using either surface measurements or satellite data show differences in the regional attribution of the global methane increase in this period. This is true also for inversions using IASI or TROPOMI data, which will be discussed in the presentation, including the added value of combining SWIR and TIR satellite data and what is needed to make optimal use of existing multi-year satellite datasets.

This work presents an assessment of the potential of WordView-3 (VW-3) satellite for the mapping of methane point source emissions at very high spatial resolution. The 3.7 m spatial resolution of WV-3 in the shortwave infrared (SWIR) combined with a high-quality noise, rich spectral configuration and pointing capabilities (daily or better revisit time), can be of great support in the global methane reduction activities to mitigate climate change. The proposed retrieval methodology is based on a multiple linear regression of six SWIR bands barely affected by methane absorption against the B7 (2235-2285 nm) that is positioned at a highly-sensitive methane absorption region. The end-to-end sensitivity analysis of the proposed methane retrieval has helped to understand retrieval errors and detection limits. The potential of WV-3 for methane mapping has been further tested under real-case studies with a positive detection of 26 different point source emissions covering different methane hotspot regions such as the O&G extraction fields in Algeria and Turkmenistan, and the Shanxi coal mining region in China. Under these real-case scenarios, we have proven the usefulness of the unique spatial resolution of the mission to even pinpoint to very small leaks (

Recent advances in satellite-based monitoring methods have enabled the detection and quantification of methane (CH4) point emissions. In particular, the combination of TROPOMI regional-scale data with high spatial resolution data from hyperspectral (high detection sensitivity with 30m pixel resolution) and multispectral (high temporal resolution and global coverage with 30-20m pixel resolution) optical missions creates an ideal framework to detect, quantify and monitor individual CH4 emissions of highly emitting point sources. In this contribution, we illustrate this synergy between the three data types to analyse the emissions from the West Coast of Turkmenistan, one of the largest methane hotspots observed by TROPOMI. In the January 2017-November 2020 period, we have found 29 individual CH4 sources directly linked to the O&G sector, of which 86% had been identified as inactive flares that vent gas. In the period of the study, the Sentinel-2 satellite has detected a total of 944 CH4 plumes with fluxes >1700kg/h, and the ZY1 and PRISMA hyperspectral satellites have quantified 25 plumes with emissions ranging from 1.400 ± 400 kg/h to 19.600 ± 8.100 kg/h. At the regional level, 2020 shows a substantial increase in the number of methane plume detections concerning previous years. Further, according to VIIRS data, the use of flaring in the country has been decreasing significantly in recent years. This suggests a causal relationship between the decrease in flaring and the increase in venting. On the other hand, the historical Landsat record shows that gas vent emissions in this region have been occurring sporadically since at least the late 80' s. This study shows that the synergy between the different satellites has been able to reveal with precision the origin of large anthropogenic CH4 emitters, which are -likely- easy to fix targets, and monitoring them over time. It has also shown the outstanding strong emissions present over the West Coast of Turkmenistan, which come from the O&G sector.

We describe a new methodology for deriving source-specific emission ratios of nitrogen oxides (NOx) to carbon dioxide (CO2) from space-based TROPOspheric Monitoring Instrument (TROPOMI) and Orbiting Carbon Observatory-2 (OCO-2) observations. The approach is based on scaling the observed ratio along the OCO-2 track with simulated data, in order to obtain the NOx-to-CO2 emission ratio at the source. We analyze fourteen TROPOMI/OCO-2 collocations from near the Matimba coal-fired power station in South Africa. We obtain a mean NOx-to-CO2emission ratio of 2.6×10-3 and standard deviation of 0.6×10-3 (or 23%). When applied to NOx emission estimates derived from TROPOMI data, we obtain annual CO2 emissions of 62, 60 and 58 kt/d for the years 2018, 2019 and 2020, respectively, with standard deviation of 20 kt/d (or 33%). These values are consistent with existing inventories such as the Open-source Data Inventory for Anthropogenic CO2 (ODIAC). The proposed method will also work ideally for new and upcoming satellite observations systems such as OCO-3, CO2M and GOSAT-GW. Reference: Janne Hakkarainen, Monika E. Szeląg, Iolanda Ialongo, Christian Retscher, Tomohiro Oda, and David Crisp: Analyzing nitrogen oxides to carbon dioxide emission ratios from space: A case study of Matimba Power Station in South Africa, Atmospheric Environment: X, Volume 10, doi:10.1016/j.aeaoa.2021.100110, 2021. https://doi.org/10.1016/j.aeaoa.2021.100110

We investigate the impact of scattering-related XCO2 uncertainties in CO2M data on inferring CO2 fluxes from cities. Our analyses focus on single overpasses over Berlin, Beijing and Shanghai under favorable conditions. To this end, we perturb synthetic CO2M data according to XCO2 uncertainty characteristics derived with a neural network approach. We assimilate them in a new flux inversion model based on an Ensemble Kalman Filter method with WRF as the transport model (available at https://git.wur.nl/ctdas). Scaling whole city fluxes yields uncertainties of 11–25% in Berlin and -1–9% in the Chinese cities. The slightly more accurate results in China may be explained by larger XCO2 enhancements due to higher emissions, while simulated XCO2 uncertainties are similar. We further attempt to constrain individual hotspots of CO2 emissions located inside of Berlin. This yields higher errors of up to 110% due to conflation of several CO2 sources and sensitivity to spatial variations of XCO2 biases on the scale of individual CO2 plumes. Our results suggest promising performance of CO2M with respect to scattering-related XCO2 uncertainties for snapshots of whole city fluxes. However, we expect larger flux uncertainties if not all favorable conditions in our hand-picked scenes are met: large emitters, cloud-free skies, low wind speeds and average aerosol load.Therefore, constraining smaller cities and annual emissions with policy-relevant accuracy requires efforts to minimize XCO2 uncertainties as well as other error sources in CO2 flux estimation methods, which are not included in our analyses.

High latitudes pose significant challenges to reliable space-based observations of total column carbon dioxide (XCO2). In addition to large solar zenith angles and frequent cloud coverage over the Arctic and boreal regions, snow-covered surfaces absorb strongly in the near-infrared wavelengths coinciding with the CO2 absorption channels of the Copernicus Anthropogenic CO2 Monitoring (CO2M) Mission. Because of the resulting low radiances of the reflection measured by the satellite, retrievals over snow may be less reliable and, for current missions, are typically filtered or flagged for potentially poor quality. Here, we present the first results from our ongoing ESA project SNOWITE. In the project, we carry out the first systematic and dedicated feasibility study of CO2 retrievals over snow. We will introduce a measurement-based snow reflectance model to be tested for simulated CO2M radiances and potential improvements in retrievals. In parallel, we analyse existing high-latitude retrievals of CO2 from the Orbiting Carbon Observatory -2. We evaluate these against the ground-based XCO2 retrievals from the Total Carbon Column Observing Network (TCCON), and identify the snow-covered satellite pixels using auxiliary snow coverage data. With the approaches in project SNOWITE, we aim to quantify the uncertainties in XCO2 retrievals over snow, and ultimately help increase the quantity and reliability of the retrievals at high latitudes in the shoulder seasons (late autumn and early spring) – an important period for the carbon cycle in the rapidly changing Arctic climate.

The Copernicus Carbon Dioxide Monitoring (CO2M) satellite mission, planned to be launched in 2025, will provide new capabilities of imaging CO2 and NO2. CO2M will retrieve XCO2 with 2×2 km2 resolution and with a precision better than 0.7 ppm, which enables a quantitative assessment of CO2 emissions at the scale of large point sources such as power plants. Yet, interpreting the data is complicated: the CO2 signal may be only weakly enhanced over the background, may be compromised by biospheric signals, and may be obscured by cloud cover because

Space-based measurements of carbon dioxide (CO2) are the backbone of the global and national-scale carbon monitoring systems that are currently being developed to support and verify greenhouse gas emission reduction measures. Current and planned satellite missions such as JAXA’s GOSAT and NASA’s OCO series and the European Copernicus Anthropogenic Carbon Dioxide Monitoring (CO2M) mission aim to constrain national and regional-scale emissions down to scales of urban agglomerations and large point sources with emissions in excess of ~10 MtCO2/year. Here, we report on the DLR-funded compact satellite mission concept “CO2Image”, which is now in Phase B. The mission will complement the suite of planned CO2 sensors by zooming in on facility-scale emissions, detecting and quantifying emissions from point sources as small as 1 MtCO2/year. A fleet of CO2Image sensors would be able to monitor roughly 80% of the CO2 emissions from coal-fired power plants worldwide. The key feature of the mission is a target region approach, covering tiles of ~50 x 50 km2 extent with a resolution of 50 x 50 m2. Thus, CO2Image will be able to resolve plumes from individual localized sources, providing higher resolution nests for survey missions such as CO2M. A demonstrator mission is planned for launch in 2026. One of the mission design questions that is currently being addressed is the optimal local overpass time to achieve the mission goals, taking into account the diurnal variability of turbulence, cloud cover, wind speed and uncertainty, and measurement geometry, and how these affect the detection of plumes and quantification of emissions. First results suggest an advantage to measuring earlier in the day, in contrast to the afternoon overpass time of most current missions targeting greenhouse gases.

Climate action plan is high on the agenda in Europe and worldwide with the goal to cut greenhouse gas emissions and reach carbon neutrality in the coming years. Continuous long-term monitoring by performing precise and accurate measurements of greenhouse gases (GHGs) and other climate relevant gases is key to measure the success of its implementation and/or identify the need for stronger actions. The nadir looking satellites, measuring the complete atmospheric column, provide the most powerful method for global mapping of these gases. However, the satellite measurements require accurate validation through high quality reference measurements. Total column concentrations of the GHGs and other climate relevant gases retrieved from ground-based solar absorption measurements using Fourier transform infrared (FTIR) spectrometers are a primary source of reference data for validating satellite data, because they sample the whole atmosphere, similar to the satellites. The Total Carbon Column Observing Network (TCCON) and the Network for the Detection of Atmospheric Composition Change – Infrared Working Group (NDACC-IRWG) have been performing solar absorption measurements using high resolution FTIRs for many years. The retrieval of total and/or partial column concentrations of GHGs and other climate relevant gases are performed from the measured spectra in the near-infrared and mid-infrared regions. These data form the baseline for validation of satellite derived trace gas products. However, the number of stations is limited and has an uneven geographical coverage. To fill in this gap, several portable low-resolution FTIRs, one of which is the EM27/SUN that is used by the Collaborative Carbon Column Observing Network (COCCON), have been extensively tested and characterized in the framework of ESA’s Fiducial Reference Measurements for Ground-Based Infrared Greenhouse Gas Observations (FRM4GHG; https://frm4ghg.aeronomie.be/) project and showed excellent performance. The spectrometers demonstrated their ability of providing high quality data comparable to that of TCCON. These low-resolution spectrometers are useful to achieve a denser distribution of the ground-based stations, cover geographical gaps for various atmospheric conditions, source regions of special interest, and to create a large latitudinal distribution of stations. In this presentation, we will outlay the state-of-the-art of the current high- and low-resolution FTIR spectrometers and present plans for future improvements. Furthermore, with the help of exemplary satellite validation cases, we will show the benefits and complementarity of using ground-based FTIR data from different spectral regions and spectral resolutions, benefits of direct comparison or considering smoothing errors.

The first satellite-based global retrievals of terrestrial sun-induced chlorophyll fluorescence (SIF) were achieved in 2011. Since then, a number of global SIF datasets with different spectral, spatial and temporal sampling characteristics have become available to the scientific community. These datasets have been useful to monitor the dynamics and productivity of a range of vegetated areas worldwide, but the coarse spatio-temporal sampling and low signal-to-noise ratio of the data hamper their application over small or fragmented ecosystems. This is being greatly alleviated by the advent of TROPOMI, which allows SIF retrievals with a much higher spatio-temporal sampling than previous missions. In this work, we present a global SIF dataset produced from TROPOMI measurements within the TROPOSIF project funded by ESA’s Sentinel-5p+ Innovation activity. This contribution will focus on SIF retrieval and validation aspects. The current version of the TROPOSIF dataset covers the time period between May 2018 and August 2021. It includes SIF retrievals from two different fitting windows. Surface reflectance in seven spectral positions within the 665–785 nm range are also included in the TROPOSIF dataset as an important ancillary variable to be used in combination with SIF. TROPOSIF data are being widely used by the scientific community already.

Abstract The Mount Etna volcanic eruption on March 12, 2021 (06:18 UTC) was followed by a north-eastward transport of the volcanic plume towards the Middle East, crossing both Antikythera/Greece (19:30 UTC) and Limassol/Cyprus (March 13, 2021, 05:30 UTC) lidar stations. According to the Volcano Observatory Notice for Aviation (VONA), reported by the Istituto Nazionale di Geofisica e Vulcanologia, Osservatorio Etneo (INGV-OE), the volcanic eruption resulted in strong ash emissions, observed at heights from 5 to 10 km a.s.l. The core idea of the present analysis is triggered by the detection of the aerosol plume by Aeolus in the Middle East the day after the eruption. Volcanic ash simulations initialized by high resolution (0.125ox0.125o) numerical outputs from the European Centre for Medium-Range Weather Forecasts (ECMWF) Integrated Forecast System (IFS), were performed with the FLEXPART-WRF (FLEXible PARTicle dispersion) model. Two different datasets of ECMWF IFS numerical outputs produced based on the same model configuration, one with (hkvt) and one without (hkvw) assimilation of Aeolus L2A wind profiles were used in our analysis. The concept of the current study has been adopted by those of the NEWTON (ImproviNg dust monitoring and forEcasting through Aeolus Wind daTa assimilatiON) project, funded by the European Space Agency (ESA), which focus on highlighting the potential improvements on short-term dust forecasts thanks to the assimilation of Aeolus wind. Therefore, our experiments here expand NEWTON research activities by justifying the improvements on forecasting and monitoring the dispersion of volcanic ash. A thorough evaluation of the different model simulations is performed against high-quality multi-wavelength Polly-XT lidar measurements, conducted at the European Aerosol Research Lidar Network (EARLINET) ACRTIS station of the PANhellenic GEophysical observatory of Antikythera (PANGEA) of the National Observatory of Athens (NOA), detecting the presence of the ETNA volcanic ash emissions over the Antikythera station on the 12th of March 2021, between 19:30 and 22:30 UTC. According to the performed analysis, it is revealed that the FLEXPART-WRF ash simulation is improved when the model is initialized with the IFS dataset relying on the hkvt experiment and is in good agreement with the Aeolus detection in the Middle East. Our results consist a first indication of a positive impact on forecasting volcanic ash plumes, attributed to the assimilation of wind profiles, thus expanding the applicability of Aeolus’ observations and the scientific achievements of the satellite mission. Acknowledgments: NEWTON team acknowledges support by ESA in the framework of Aeolus+Innovation (ESA AO/1-9544/20/I/NS). This research was also supported by the European Research Council (ERC) under the European Community’s Horizon 2020 research and innovation framework programme – ERC grant agreement no. 725698 (D-TECT). We acknowledge the support by EU H2020 E-shape project (Grant Agreement n. 820852). Also, this research was supported by data and services obtained from the PANhellenic Geophysical Observatory of Antikythera (PANGEA) of the National Observatory of Athens (NOA), Greece, and we acknowledge EARLINET for providing aerosol lidar profiles (https://www.earlinet.org). NOA team acknowledges the support of the Stavros Niarchos Foundation (SNF).

The representation of stratospheric winds in numerical weather prediction systems have significant uncertainties. Due to the very limited number of global stratospheric wind observations, it is difficult to validate these numerical model predictions. ESA’s Doppler Wind Lidar mission Aeolus, launched on 22 August 2018, provides accurate measurements of the horizontal line-of-sight (HLOS) wind speed from the surface up to 30 km. Aeolus wind observations are being operationally assimilated by ECMWF, Météo-France, the German Weather Service (DWD), the UK Met Office and the National Centre for Medium Range Weather Forecasting (NCMRWF). Additional to these weather prediction models, independent and high-quality reference measurements are essential for the quantification Aeolus wind biases. The Loon balloon network provides unique wind observations in the upper troposphere and lower stratosphere, which can be used to validate Aeolus winds and to better characterize atmospheric dynamics in an altitude region, where only very few in situ wind observations are available. Between 2018 and the beginning of 2021, Loon has continuously launched super-pressure balloons in the higher atmosphere, which offer an enormous reference dataset for the Aeolus product quality assessment. While Loon’s main objective was to provide internet connection to remote areas in the world, the balloons also carry sensors to measure GPS location, altitude and atmospheric variables. This study aims at the validation of Aeolus observations in the tropical lower stratosphere against Loon balloon observations. While Aeolus measures instantaneous vertical profiles of the HLOS wind speed, Loon yields temporally highly resolved and very accurate wind vectors in a specific altitude range along the balloon trajectories. Hence, Loon observations can be used to complement Aeolus wind profiles in order to characterize the variability of the horizontal wind speed. In this paper, Aeolus winds are compared against collocated Loon observations and ECMWF model winds for the years 2019 and 2020, which allows the analysis of Aeolus wind bias trends. Special emphasis is given to the Central Andes region in Latin America, where most of the Loon observations are concentrated for the considered period. This region is characterized by strong vertical wind shear and orographic effects. The potential of Aeolus to observe these effects is shown and discussed on the basis of a case study.

ESA’s Aeolus satellite observations are expected to have the biggest impact for the improvement of numerical weather prediction in the Tropics. An especially important case relating to the evolution, dynamics, and predictability of tropical weather systems is the outflow of Saharan dust, its interaction with cloud microphysics and impact on the development of tropical storms over the Atlantic Ocean. The Atlantic Ocean off the coast of West Africa and the eastern Caribbean uniquely allows the study of the Saharan Aerosol layer, African Easterly Waves and Jet, Tropical Easterly Jet, as well as the deep convection in the Intertropical Convergence Zone and their relation to the formation of convective systems, and the long-range transport of dust and its impact on air quality. The Joint Aeolus Tropical Atlantic Campaign (JATAC) deployed on Cabo Verde and the US Virgin Islands is addressing the validation and preparation of the ESA missions Aeolus, EarthCARE and WIVERN, as well as supporting the related science objectives raised above. The JATAC campaign started in July 2021 with the deployment of ground-based instruments at the Ocean Science Center Mindelo (OSCM, Cabo Verde), including the EVE lidar, the PollyXT lidar, a W-band Doppler cloud radar and a sunphotometer. By mid-August, the CPEX-AW campaign started their operations from the US Virgin Islands with NASA’s DC-8 flying laboratory in the Western Tropical Atlantic and Caribbean with the Doppler Aerosol Wind Lidar (DAWN), Airborne Precipitation and Cloud Radar (APR-3), the Water Vapor DIAL and HSRL (HALO), a microwave sounder (HAMSR) and dropsondes. In September, a European aircraft fleet was deployed to Sal (Cabo Verde) with the DLR Falcon-20 carrying the Aeolus Airborne Demonstrator (A2D) and the 2-µm Doppler wind lidar, and the Safire Falcon-20 carrying the high-spectral-resolution Doppler lidar (LNG), the RASTA Doppler cloud radar, in-situ cloud and aerosol instruments among others. The Aerovizija Advantic WT-10 light aircraft with filter-photometers and nephelometers for in-situ aerosol characterisation was operating in close coordination with the ground-based observations from Mindelo. More than 35 flights of the four aircraft were performed. 17 Aeolus orbits were underflown, four of which completed by simultaneous observations of three aircraft, with a perfect collocation of Aeolus and the ground-based observation for two cases. Several flights by the NASA DC-8 and the Safire Falcon-20 have been dedicated to cloud microphysics and dust events. The EVE lidar has been operating on a regular basis, while the PollyXT and several other ground-based instruments were continuously operating during the campaign period. For further characterisation of the atmosphere, radiosondes were launched up to twice daily from Sal airport. Additionally, there were radiosonde launches from western Puerto Rico and northern St Croix, US Virgin Islands. The JATAC was supported by dedicated numerical weather and dust simulations supporting the forecasting efforts needed for successful planning of the flights and addressing open science questions. While the airborne activities were completed end September, the ground-based observations are continuing into 2022. The paper will present an overview and initial results of JATAC. In memory of our colleague and friend Gail.

For a better comprehension of atmospheric dynamics, it is fundamentally important how well we understand the general condition (dynamics and chemistry) in the atmosphere. Aeolus wind measurements allow the derivation of atmospheric wave structures on different temporal and spatial scales and wind gradients in particular above the oceans, where wind measurements from ground-based instruments are sparse. Planetary waves (PWs) are global scale waves, which are well-known as main drivers of the large-scale weather patterns in mid-latitudes on time scales from several days up to weeks in the troposphere. When PWs break, they often cut pressure cells off the jet stream. A specific example are so-called streamer events, which occur predominantly in the mid- and high-latitudes of the lower stratosphere. Streamers are characterized by ozone-poor airmasses occurring mainly in the Northern Atlantic / European section and leading to various consequences due to a strong increase of UV radiation. At the flanks of the streamers gravity waves can be excited due to stron wind shear. We use Aeolus wind data to derivate the PW-activity. We derive the so-called dynamical activity index (DAI) based on Aeolus L2B wind measurements (9500 – 10000 m) and ERA5 reanalysis wind data (250 hPa). A first comparison of the DAI based on the two data sets is presented. First preliminary case studies are shown addressing the structure of streamers and their relation to planetary wave breaking based on Aeolus wind measurements. First results of the activity of gravity waves with respect to streamers are also shown.

For a better understanding of atmospheric dynamics, it is very important to know the general condition (dynamics and chemistry) in the atmosphere. Aeolus wind measurements provide wind measurements from satellite instrument. ERA 5 can produce very detailed information about dynamics without gaps in time series in high resolution (0.25°). Planetary waves (PWs) are global scale waves, which are well-known as main drivers of the large-scale weather patterns in mid-latitudes on time scales from several days up to weeks in the troposphere. When PWs break, they often cut pressure cells off the jet stream. A specific example are so-called streamer events, which occur predominantly in the mid- and high-latitudes of the lower stratosphere. Streamers are characterized by ozone-poor airmasses occuring mainly in the Northern Atlantic / European section and leading to various consequences due to a strong increase of UV radiation. We compare ERA5 reanalysis with Aeolus measurements. This comparison can bring us an answer if we can use ERA5 instead of Aeolus measurements in case of time gaps. We also use homogeneity test for ERA5 time series. Moreover, we also analyze characteristics of gravity waves (GW) in the ionosphere using continuous Doppler sounding and in the troposphere using large aperture array of microbarometers. Similarly, ground based infrasound monitoring is performed. We investigate, if there are any changes of GW or infrasound characteristics related to stratospheric processes, e.g., streamer events.

Dust aerosols consist a key component of the Earth-Atmosphere system with multifarious impacts on atmospheric processes, ecosystems, humans’ health, air quality and various socio-economic sectors. Based on this fact, it is clearly reflected the imperative need to monitor and predict the characteristics of dust burden by employing optimally highly accurate observations and state-of-the-art atmospheric-dust models. This is the overarching goal of the NEWTON (ImproviNg dust monitoring and forEcasting through Aeolus Wind daTa assimilatiON) project, funded by the European Space Agency (ESA), under the framework of the Aeolus+Innovation call. NEWTON research activities are performed via the collaboration of the National Observatory of Athens (NOA), the Cyprus Institute (CyI) and the European Centre for Medium range Weather Forecasts (ECMWF). The core idea of NEWTON triggered by the remarkable improvements on numerical weather prediction, attributed to the assimilation of Aeolus wind profiles, as it has been evidenced in several weather centers worldwide and recent studies. Such advancements are expected to benefit subsequently dust numerical simulations since mineral particles’ life cycle is governed mainly by the prevailing atmospheric circulation. For addressing the NEWTON scientific objectives, regional dust numerical simulations along with spaceborne and ground-based aerosol observations, acquired by active and passive remote sensing instruments, are utilized. Two versions of the WRF-Chem model, operating at NOA and CyI, are initialized with ECMWF IFS outputs produced with (hel4) and without (hel1) the assimilation of Aeolus quality-assured Rayleigh-clear and Mie-cloudy wind profiles. Our experiments are performed for two regions of interest (ROIs), the Tropical Atlantic Ocean-Western Sahara and the broader Eastern Mediterranean, subjected frequently to dust transport, encompassing also the major natural erodible dust sources of the planet. For the justification of the anticipated improvements on short-term dust forecasts, relying on the hel4 experiment, the dust-related numerical products are evaluated against ground-based columnar and vertically resolved aerosol optical properties acquired by AERONET sunphotometers and PollyXT lidars operating in the two ROIs. Moreover, for evaluation purposes, the wealth of data from the EARLINET Covid-19 and ASKOS experimental campaigns are exploited. Our assessment analysis is upgraded by including the NOA in-house LIVAS and MIDAS satellite datasets providing vertical and columnar dust optical properties, respectively. According to the performed analysis, it is revealed that for several cases, either of high or low aerosol conditions, the WRF predictive skills are improved when the regional simulations are initialized after the consideration of Aeolus wind assimilation (hel4). Nevertheless, this behaviour is not consistent in space and time since in some cases the two WRF runs perform equally and in a few cases Aeolus wind assimilation downgrades models’ performance. Overall, there are signals of positive impacts of Aeolus wind assimilation on dust forecasting, which, however, their validity must be further ensured. In addition, an in-depth investigation of how Aeolus winds affect the simulated atmospheric state and particularly the meteorological factors driving dust emission, transport and removal, is on-going. Finally, NEWTON progress, activities and achievements are disseminated via the official website (https://newton.space.noa.gr) and the EO4Society portal (https://eo4society.esa.int/).

The Atmospheric Laser Doppler Instrument (ALADIN) is the world’s first space-based Doppler wind lidar. ALADIN operates at 355nm and its main products are line-of-sight wind profiles derived from the direct detection of the Doppler shift of the backscattered signals. Using a variation of the High Spectral Resolution Lidar technique (HSRL), two main detection channels are used, a `Mie ‘-channel and a `Rayleigh’-channel. ALADIN’s design is optimized for wind observations, however, Cloud and aerosol information can also be retrieved from the attenuated backscatter signals. ATLID (Atmospheric Lidar) is the lidar to be embarked on the Earth Clouds and Radiation Explorer (EarthCARE) mission. EarthCARE is a joint ESA-JAXA mission and will embark a cloud/aerosol lidar (ATLID) as well as a cloud-profiling Radar (CPR) a multispectral imager (MSI) and a three—view broad-band LW and SW radiometer (BBR). ATLID is a HSRL systems, however, unlike ALADIN, ATLID does not measure winds and is optimized exclusively for cloud and aerosol observations. In particular, compared to ALADIN, ATLID has a higher spatial resolution, measures the depolarization of the return signal and has a higher contrast molecular Rayleigh vs cloud/aerosol backscatter signal separation. In spite of their differences, both ALADIN and ATLID face similar challenges when it comes to the retrieval of aerosol and cloud properties. The most important challange is the fact that the SNR ratio of the backscatter signals is low compared to e.g. those associate with terrestrial HSRL or Raman lidars. The low SNR of the atmospheric signals creates difficulties when using standard HSRL inversion methods. Along-track averaging can increase the SNR, however, the presence of clouds (and inhomogeneities in general) may lead to large biases in the retrieved extinction and backscatters if not accounted for. Cloud/aerosol algorithms have been developed for ATLID that have focused on the challenge of making accurate retrievals of cloud and aerosol extinction and backscatter specifically addressing the low SNR nature of the lidar signals and the need for intelligent binning/averaging of the data. Two of these processors are A-FM (ATLID featuremask) and A-PRO (ATLID profile processor) A-FM uses techniques inspired from the field of image processing to detect the presence of targets at high resolution, while A-PRO (using A-FM as input) impliments a multi-scale optimal-estimation approach in order to retrieve both aerosol and cloud extinction and backscatter profiles. Versions of the A-FM and A-PRO processors have been developed for Aeolus (called AEL-FM and AEL-PRO, respectively). Prototype codes exist and are in the process of being introduced into the L2a operational processor. In this presentation, AEL-FM and AEL-PRO will be described and representative results presented and discussed. Advances in creating accurate cross-talk corrected signals using the ALADIN Mie channel will also be discussed.

Global aerosol monitoring infrastructure is regularly complemented by new instrumentation to deepen our understanding of aerosols, which are key agents of the global radiative budget. Here, we report the first results from an AOD (Aerosol Optical Depth) product based on the combination of lidar surface returns from Aeolus and collocated near surface wind speed estimates over oceans. The retrieval is based on the fundamental link between lidar surface returns, sea slope variance and near-surface wind over water. We apply state-of-the-art parametrization of sea surface reflectance optimized for non-nadir incidences (required for Aeolus which has an incidence angle of 37.5o) which has not been empirically applied to any spaceborne system to date. Firstly, we evaluate the used method to CALIPSO (The Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation) lidar surface return, yielding AOD estimates in good agreement with the existing surface-based estimates of AOD from the SODA algorithm (synergy between CALIPSO and CloudSat surface returns). Second, we describe how the used parametrization can be exploited for Aeolus retrievals given its unique characteristics and low signal-to-noise ratios. Finally, the retrieved AOD from the surface return is compared to the collocated L2a Aeolus operational extinction profile retrievals for selected days of the Aeolus-IOP (Intensive Observation Period in September 2019). In the presentation we will show the first promising results of this new retrieval algorithm. The final AOD retrieval will be beneficial from both research and operational perspectives. In particular, it can (a) elucidate the usefulness of lidar surface returns from Aeolus-like systems (355 nm and non-nadir incidence for AOD or surface wind speed retrievals over sea), (b) extend the planned operational algorithms of future UV (Ultraviolet) lidar missions (EarthCARE and Aeolus Follow On) with AOD retrieval independent from aerosol microphysics assumptions.

ADM-AEOLUS is a satellite from the European Space Agency and has been launch on 22 August 2018. The satelliteis designed to measure wind vertical profiles along its trace all over the planet. To fulfill this mission it carries a High Spectral Resolution Lidar (HSRL) : Atmospheric LAser Doppler Instrument (ALADIN). The lidar emit light in the UV wavelength (λ=354,8 nm). Then the lidar signal is backscattered by the atmosphere and is collected by twodetection channels: The Rayleigh channel more sensitive to the molecular backscatter and the Mie channel more sensitive to the particle backscatter. The detected signal is accumulated from 0 to 25 km in bins that range between 250 m and 2km. The satellite take one measurement every ≈3 km along its track, which are usually accumulated by group of 30 in Basic Reapeat Cycle (BRC) that is 87 km long. This is the accumulation length that is processed by the algorithms presented hereafter. The paper will present the two algorithms of the Level 2A processor that aims to retrieve the atmospheric optical properties from the ADM-AEOLUS signal. The first algorithm is the Standard Correct Algorithm (SCA). The SCA does a direct retrieval of extinction and backscatter coefficients. The optical properties of the molecular atmosphere can be simulated from pressure and temperature profiles. Lidar signals are processed to separate the contributions from molecules and particles in a step called cross-talk correction. The optical properties of the particles can be estimated from these cross-talk corrected signals and simulated molecular signals. For the extinction this is done iteratively to account for the attenuation of the overlying layers. It results that the extinction is particularly sensitive to the noise as an error in each bin will propagate below. The SCA also compute the coefficients for the middle bins by averaging two adjacent bins which reduce the errors and provides more reliable extinction coefficients. The second algorithm called Maximum Likelihood Estimation (MLE) is being implemented. The MLE determine a profile that optimally fits the signals while still checking physical constraints: The extinction is not allowed to be negative and the lidar ratio is bounded between 2 and 200 sr. By doing so, the results are less sensitive to the noise than the SCA and, in particular, it improves the estimation of the extinction coefficients. Finally, examples from real ADM-AEOLUS observations will be shown to illustrate the performances of those algorithms.

In recent years, aerosol prediction has become more skillful thanks also to advances in aerosol observability. Several centres with forecasting capabilities now offer forecasts of aerosol fields up to five days ahead. At ECMWF, the Copernicus Atmospheric Monitoring Service (CAMS) delivers aerosol products based on analysis of Aerosol Optical Depth (AOD) from various sensors. Recent reasearch efforts have also shown the potential of lidar backscatter to extract aerosol information, providing also important insight on the aerosol vertical structure. In the frame of the Aeolus mission, ESA funded a collaboration between several institute to work on the operational phase (DISC). The Aeolus Aerosol Assimilation in the DISC (A3D) has the mission to look at the potential of lidar data from the ALADIN lidar instrument on boad of Aeolus. While this instrument is designed for wind retrieval and not specifically for aerosol science, some useful information can be extracted through a combination of post-processing and assimilation. This is also preparatory work for the EarthCARE mission which will operate a UV lidar at the wavelength of 355nm, designed to study aerosols and clouds. The A3S project (precursor of A3D) demonstrated the technical feasibility of assimilating lidar backscatter signal at that frequency using demonstration datasets built on model and CALIPSO data. Following the successfull launch of Aeolus in 2018, ESA has funded a continuation of the A3S project with the goal to assimilate in-orbit data of particle backscatter (L2A) from the ALADIN instruments. In this presentation, preliminary results from this activity will be discussed, focusing in particular on the post-processing of the L2A particle backscatter product using a model-dependent cloud screening to ensure that only aerosol signal is allowed into the 4D-Var analysis.

In ESA Aerosol_cci, an ensemble algorithm had been developed for AATSR, which uses the pixel-level uncertainties in the aerosol products to combine results of several algorithms for the same sensor. This algorithm was now applied within the Copernicus Climate Change Service (C3S) to several sensors: AATSR (and ATSR-2), SLSTR, OLCI (and MERIS) and IASI. Before calculating the ensemble results, the pixel-level uncertainties of all algorithms / instruments were evaluated against true errors versus AERONET applying new methodology developed in Aerosol_cci+ and the H2020 project FIDUCEO. This evaluation included the comparison of histograms of true errors with integrated error distributions calculated by super-positions of Gaussian error distributions with the width of the pixel-level uncertainties in the products. Secondly, percentiles in pixel-level uncertainties were assessed as function of binned true errors to judge, how well the pixel-level uncertainties can discriminate "good" and "bad" pixels. Those evaluations proved the high performance of part of the pixel-level uncertainties, while they also revealed some weaknesses of others. Based on these evaluations, piece-wise linear corrections of those pixel-level uncertainties were manually developed and applied, before the ensemble values were calculated. Validation of the gridded daily ensemble products with the AEROCOM tools showed clearest beneficial effects for the IASI ensemble (improved coverage, 4 algorithms contribute), slight effects for (A)ATSR(-2) and SLSTR ( smallest bias, best consistency of the two ATSR sensors; 3 algorithms contribute) and no benefit for OLCI (only two algorithms involved). The paper will show selected results of all those aspects.

In the 2020s, a number of satellites with Multi-Angle Polarimetric (MAP) instruments will be launched, such as METOP-SG from ESA/EUMETSAT with the 3MI polarimeter, the NASA PACE mission with onboard the SPEXone and HARP-2 polarimeters, the ESA CO2M-mission, and in the late 2020s the NASA ATMOS mission. MAP measurements provide the highest information content on aerosol properties from a passive remote sensing point of view, and allow accurate retrieval of aerosol optical properties (optical thickness, single scattering albedo, phase function) and microphysical properties (size distribution, refractive index, shape), needed for climate and air quality research. So, the expectation is that the quality of aerosol remote sensing measurements will advance significantly in the coming years. To cope with the increased information content of MAP instrumentation advanced retrieval algorithms need to be (further) developed. Here, full inversion approaches are needed that consider a continuous space of aerosol microphysical properties (size distribution, refractive index), instead of using standard aerosol models, and to properly account for land or ocean reflection by retrieving land or ocean parameters simultaneously with aerosol properties. Currently, there are two full inversion algorithms that have proven capability at a global scale: the Generalized Retrieval of Aerosol and Surface Properties (GRASP) algorithm, developed at University of Lille and the GRASP-SAS company, and the Remote Sensing of Trace gas and Aerosol Products (RemoTAP) algorithm, developed at SRON Netherlands Institute for Space Research. Currently, the ESA HARPOL (Harmonizing and advancing retrieval approaches for present and future polarimetric space-borne atmospheric missions) project is ongoing with the following objectives: 1) Identify strong/weak points for both GRASP and RemoTAP through intercomparison of aerosol/surface products and harmonized validation. 2) Define the optimal set of aerosol properties to be retrieved from MAP measurements. 3) Define the optimal aerosol, land surface and ocean reflection models to be used in MAP retrievals. 4) Define the (possible) need for prior information. 5) Understand the intrinsic limitations of aerosol retrieval from past and upcoming MAP measurements. 6) Provide recommendations for future space born missions dedicated to atmospheric aerosol studies. Here, we report on the first results from the HARPOL project, based on both real measurements from POLDER-3/PARASOL as well as synthetic measurements.

Aerosols affect climate in several ways. Aerosols together with clouds contribute the largest uncertainties to the Earth’s energy budget, according to IPCC. Consequently, accurate retrieval of the Aerosol Optical Depth (AOD) from satellite measurements is important to get more knowledge about aerosols in the atmosphere and the influence of natural and anthropogenic events on the amount of aerosols. Since the retrieval of AOD needs assumptions concerning aerosol properties and the surface of the Earth there are several different algorithms. We analyse data from the Copernicus Climate Change Service of retrieved AOD with Dual-View Instruments (Along Track Scanning Radiometer 2 (ATSR2), Advanced Along Track Scanning Radiometer (AATSR), Sea and Land Surface Temperature Radiometer (SLSTR)) and the Infrared Atmospheric Sounding Interferometer (IASI) for the retrieval of Dust AOD. For reliable conclusions there should be a consistency of these algorithms and between the different instruments. In a comparison of AOD trends in different regions we analyse this consistency and the comparability of the different instruments. The trends were calculated by a seasonal Mann-Kendall-trend-test after a bias correction between the different instruments. For further validation we compare the trends to AERONET ground based measurements. Here the distribution of the AERONET stations in the considered region can cause problems, for example when they are not representative leading to opposite trend between AERONET and satellite observations for example in Asia and North Africa. Overall, we find good consistency for most regions with few comparisons revealing larger inconsistencies. This will be further explored in the presentation.

The strong economic development in China, which started by the end of the previous century, and the associated urbanization and socio-economic development, have led to the increase in the emissions of aerosols and trace gases, including aerosol pre-cursors such as SO2, NO2 and volatile organic compounds (VOCs). These increases were clearly observed in satellite observations of AOD (e.g. de Leeuw et al., 2018; Sogacheva et al. 2018) and total column vertical densities (TVCDs) of, e.g. SO2 and NO2 (e.g. Van der A er al. 2017), together with the decline of such components in response to emission reductions in response to the implementation of a series of clean air action plans (Van de A et al., 2017; Sogacheva et al., 2018). However, in a recent study, using OMI and TROPOMI data, we observed that the downward trends in NO2 TVCDs were reversed in recent years, with different behavior in different parts of China, roughly divided by the Yangtze River (Fan et al., 2021).This observation inspired us to investigate the behavior of the AOD and preliminary results indicate that also the AOD trend has reversed in recent years and AOD does not decrease as much as during the last decade, or has even increased recently (de Leeuw et al., 2021). Detailed results will be presented including trends over the densely populated provinces in China. References­ -de Leeuw, G., Sogacheva, L., Rodriguez, E., Kourtidis, K., Georgoulias, A. K., Alexandri, G., Amiridis, V., Proestakis, E., Marinou, E., Xue, Y., and van der A, R.: Two decades of satellite observations of AOD over mainland China using ATSR-2, AATSR and MODIS/Terra: data set evaluation and large-scale patterns, Atmos. Chem. Phys., 18, 1573-1592, https://doi.org/10.5194/acp-18-1573-2018, 2018.­ -de Leeuw, G., R. van der A, J. Bai, Y. Xue, C. Varotsos, Z. Li, C. Fan, X. Chen, I. Christodoulakis, J. Ding, X. Hou, G. Kouremadas, D. Li, J. Wang, M. Zara, K. Zhang, Y. Zhang (2021). Air Quality over China. Remote Sens. 2021, 13, 3542. https://doi.org/ 10.3390/rs13173542.­ -Fan, C., Li, Z., Li, Y., Dong, J., van der A, R., and de Leeuw, G.: Variability of NO2 concentrations over China and effect on air quality derived from satellite and ground-based observations, Atmos. Chem. Phys., 21, 7723–7748, https://doi.org/10.5194/acp-21-7723-2021, 2021.­ -Sogacheva, L., Rodriguez, E., Kolmonen, P., Virtanen, T. H., Saponaro, G., de Leeuw, G., Georgoulias, A. K., Alexandri, G., Kourtidis, K., and van der A, R. J.: Spatial and seasonal variations of aerosols over China from two decades of multi-satellite observations – Part 2: AOD time series for 1995–2017 combined from ATSR ADV and MODIS C6.1 and AOD tendency estimations, Atmos. Chem. Phys., 18, 16631-16652, https://doi.org/10.5194/acp-18-16631-2018, 2018.­ -van der A, R. J., Mijling, B., Ding, J., Koukouli, M. E., Liu, F., Li, Q., Mao, H., and Theys, N.: Cleaning up the air: effectiveness of air quality policy for SO2 and NOx emissions in China, Atmos. Chem. Phys., 17, 1775–1789, https://doi.org/10.5194/acp17-1775-2017, 2017

Stratospheric aerosols play an important role in the Earth system and in the climate. Through the scattering of solar radiation back to space and by heating the stratosphere through the absorption of thermal infrared radiation upwelling from the troposphere, stratospheric aerosols impact the radiative forcing and thus the energy balance of the Earth’s atmosphere. By providing a surface for heterogeneous reactions, which release halogens, stratospheric aerosols contribute to the catalytic depletion of ozone. A strong coupling between the stratospheric aerosols, stratospheric ozone amount, and thermal balance and dynamics of the atmosphere, determines the need of global long-term aerosol records for climate modelling studies and interpretation of the long-time measurement time series related to the stratosphere. So far, only a few attempts were made to merge stratospheric aerosol records from different instruments. The most comprehensive outcome is the version 2.0 of the Global Space-based Stratospheric Aerosol Climatology (GloSSAC) comprising merged data from 10 instruments covering the period from 1978 to 2018. The main focus of GloSSAC is the widest time range and global coverage, while the number of instruments included in the record after the year 2000 (OSIRIS, CALIPSO, SAGE III/ISS) is kept at minimum. The main objective of a joint study of the Finnish Meteorological Institute and the University of Bremen is to elaborate an alternative long-term aerosol record with increased reliability by including multiple instruments measuring similar atmospheric quantities in the post-SAGE II period. This is planned to be done by including measurements from GOMOS, SCIAMACHY and OMPS-LP instruments in addition to those from SAGE II, OSIRIS, and SAGE III/ISS into the merged time series. The study begins with an analysis of differences and potential drifts between the time series of the stratospheric aerosol extinction coefficients from different instruments. Furthermore, errors associated with the conversion of the aerosol extinction coefficient between different wavelengths are analyzed. The results of these initial activities will be reported in the presentation.

S5p/TROPOMI measurements in the UV provide unique information about absorption and elevation properties of aerosols. Moreover, measurements in the wide spectral range are very sensitive to aerosol size and surface type. In the framework of the ESA S5p+Innovation AOD/BRDF project an innovative algorithm for aerosol and surface retrieval from the S5p/TROPOMI instrument has been developed. It integrates the advanced GRASP algorithm, the heritage aerosol optical depth (AOD) and directional Lambertian equivalent reflectivity (DLER) algorithms. The GRASP algorithm has already been successfully applied to space-borne instruments like PARASOL, MERIS, OLCI for simultaneous retrieval of aerosol and surface properties in the VIS and SWIR spectral range. The heritage algorithm has been previously applied to TOMS, GOME(-2), SCIAMACHY and OMI sensors for AOD and DLER characterization. Together with basic aerosol and surface characteristics like AOD and different surface albedos, the innovative algorithm for S5p/TROPOMI instrument provides extended aerosol characterization from UV to SWIR spectral range including the Angstrom exponent, spectral aerosol single scattering albedo, surface BRDF and DLER. The results of aerosol and surface validation and inter-comparison obtained in the frame of the ESA S5p+Innovation AOD/BRDF project will be presented here and the new advanced possibilities of aerosol and surface characterization from S5p/TROPOMI instrument will be discussed.

S5P/TROPOMI was launched on 13 October 2017 and provides a suite of products for the characterisation and monitoring of atmospheric trace gases, clouds, and aerosols. TROPOMI provides spectral information from the UV, VIS, NIR, and SWIR at a moderately high spectral resolution, as a successor of OMI, GOME(-2) and SCIAMACHY. The aerosol optical thickness (AOT) and surface Lambertian Equivalent Reflectivity (LER) products that were developed for these missions were ported to TROPOMI within the ESA S5P+Innovation AOD/BRDF project, in order to provide increased knowledge of the surface characterisation and atmospheric correction to end-users and L2 products developers alike. Additionally, a directional LER (DLER) product was developed. This new product wil provide new essential information about the surface chracterisation at the TROPOMI geometry. Within the project, the above products were integrated as ancillary information into the advanced GRASP algorithm, which provides state-of-the-art aerosol and surface characterisation from the UV to SWIR, including Ängström exponent, spectral single scattering albedo and surface BRDF and DLER. THE AOT and (D)LER products were developed in parallel, making use of the same Look-Up table definitions, cloud filtering, and reflectance preprocessing. 21 spectral channels were defined at the most optimal wavelengths throughout the UV, VIS and NIR, avoiding gaseous absorption, which were input to the LER product. A subset of five wavelengths in the UV were used for the AOT, while the LER product is input to the DLER product. This allowed a very efficient, but computationally demanding, processing. In order to facilitate this, ESA provided the Product Algorithm Laboratory (PAL), which was developed to provide technological support to algorithm development, and a smooth transition to operationally produced products. We will highlight our experiences with PAL. The newly developed (D)LER and AOT products will be discussed, showing the advances compared to predecessor missions and their accuracies.

Several studies have shown that aerosol retrieval from satellites is strongly affected by cloud contamination errors and cloud enhancement. In the transition zone between clouds and cloud-free air, cloud enhancement leads to an increase of aerosol optical thickness and to changes in the aerosol particle size. The choice of the cloud mask to be used in aerosol retrieval (AOT) applications is thus critical. The ESA Aerosol-CCI project showed the effect of using a common cloud mask to different aerosol retrieval algorithm as well as the impact of increases the “safety zone” around clouds to reduce possible enhancement effects (Holzer-Popp et al., 2013). In the Aerosol-CCI+ project a new innovative algorithm, CISAR, is applied to S3A/SLSTR observations. The CISAR algorithm (Govaerts and Luffarelli, 2018; Luffarelli and Govaerts, 2019) has been extended to the retrieval of cloud optical properties to overcome the need of an external cloud mask. After a so-called training period, the CISAR algorithm process all available satellite observations, i.e. both cloudy and cloud free air. The new CISAR version, developed in the framework of the ESA SEOM CIRCAS project, has been applied to S3A/SLSTR observations aggregated at 5 km. Aerosols are retrieved in the vicinity of clouds as well as within optically thin clouds, assuring a larger spatial coverage than traditional aerosol retrieval algorithm. The new CISAR version also helps reducing the overestimation at low AOTs. CISAR faces many challenges, such as discriminating dust events from clouds, surface reflectance contamination due to extended cloud coverage and the discrimination between aerosol and clouds at low optical thicknesses. Figure 5 Jacobian time series related to one year of SLSTR acquisition over Carpentras, France (top panel), and Fowlers Gap, Australia (botton panel). These challenges are addressed in the framework of Aerosol-CCI+ by implementing spectral constraints on both the surface model parameters and the aerosol and cloud single scattering properties. CISAR is then applied to S3A/SLSTR observations to obtain a 1-year global product at 10km spatial resolution.

Satellite- and ground-based measurements of sulfur dioxide emissions can obtain different flux results. When comparing satellite results with ground measurements for the 2020/21 Kilauea eruption, the same SO2 emission trend was visible; high emissions during the first week, followed by a sharp decrease and a plateau until the end of the eruption. But the satellite flux was up to 26 times lower than what the ground measured. This research aims at finding the source of this discrepancy. The satellite measurements conducted by the TROPOMI spectrometer onboard Sentinel-5P was able to give daily SO2 measurements from Kilauea using the analysis toolkit PlumeTraj. The ground-based measurements were conducted by the USGS at the Hawaii Volcano Observatory, using a UV spectrometer mounted on a car. It was found that the flux was calculated differently; the two methods used different wind speeds, the remote sensing one was measured by the Global Data Assimilation System and the other with an anemometer. New calculations were done in order that the two measurements used the same wind speeds. It was concluded that when both use the same wind speeds, the fluxes are similar when taking into consideration the standard deviation. This means that wind speed influences flux calculations. On this basis, wind speed accuracy is primordial to create a reliable flux time series.

After approximately four months of effusive activity, La Soufrière volcano (on the island of St Vincent in the Lesser Antilles Arc) erupted explosively on 9 – 22 April 2021, producing the largest emission of SO2 from Caribbean volcanoes in the satellite era. This SO2 was injected at altitudes up to 16 km a.s.l. and transported around the globe. Explosive eruptions are extremely hazardous to human health and infrastructure and can impact local or global climate through the emission of reactive gases, aerosols, and aerosol precursors, especially if they are injected into the stratosphere. For these reasons it is important to investigate the processes that drive such eruptions to better prepare for future activity and understand the impacts they may have. A vital aspect of this work is monitoring volcanic gas emissions, as changes in the flux or composition of the emitted gases can reflect changes in the magmatic system. However, determining the emission time series during an explosive eruption is extremely difficult. Ground-based measurements are not able to accurately quantify emissions due to the high altitude of emissions and the presence of volcanic ash in the plume, while satellite measurements alone can lack the spatial or temporal resolution necessary to reconstruct the emission history. Here, we tackle this problem by combining observations of SO2 measured by the TROPOspheric Monitoring Instrument (TROPOMI) onboard the European Space Agency’s Sentinel-5P satellite with PlumeTraj, a back-trajectory analysis toolkit, to reconstruct the hourly emission time-series during the onset of the explosive activity at La Soufrière. Our analysis shows that the initial explosion was gas poor, followed by a phase of continuous explosions which were gas rich (with fluxes of approximately 25 times the initial explosion). This suggests that the initial explosion cleared magma degassed during the effusive activity that occurred in the months prior to the eruption, followed by the efficient eruption of undegassed magma in the main phase of the eruption. Our results highlight the ability of satellite remote sensing for investigating volcanic processes when combined with other analysis tools. This is especially the case when estimations of gas emissions cannot be measured from the ground, such as during explosive eruptions. These results will help to improve our understanding of the physical processes driving explosive volcanism and inform future risk management strategies and monitoring at St Vincent volcano and beyond.

Nyiragongo is one of the six volcanoes that has a persistent lava lake that causes continuous SO2 emissions into the atmosphere, even during non-eruptive periods.  Nyiragongo is in the Democratic Republic of Congo near the border with Rwanda. Several cities are located around the volcano, making it a geological hazard and a risk for the surrounding population. The last recorded eruption was in 2002, when the lava flowed into the capital city of Goma, causing hundreds of fatalities. The remote location and the geopolitical situation in the region make this volcano the perfect target for monitoring by remote sensing methods. On 22 May 2021, an eruption started with the opening of a least one fissure on the volcano's flank which caused the lava lake to drain. The lava spilled over the SE flank of the volcano and flowed towards Goma where it came to a stop right before the entrance. The eruption caused approximately a hundred fatalities and the displacement of approximately 500,000 people. The Sentinel-1 mission comprised two satellite platforms with a C-Band instrument. Sentinel-1A was launched in April 2014 and Sentinel-1B in April 2016. The re-visit time over the region of interest is 6 to 12 days, depending on the orbit. The Sentinel-5P satellite platform, carrying the UV spectrometer TROPOMI (Tropospheric Monitoring Instrument), was launched on 13th October 2017 with scientific data available from May 2018.   The sub daily SO2 flux data is calculated using the analysis toolkit PlumeTraj, which allows the height and age of the SO2 emissions to be obtained. The interferograms are produced using the LiCSAR(Looking Into Continents from Space withSynthetic  Aperture  Radar) processor for both orbits. The time series of the deformation are obtained using LiCSBAS or STAMPS (Stanford Method for Persistent Scatterers). In this work, we used daily TROPOMI Level 2 SO2 data to analyse the changes in the degassing behaviour in the periods before, during and after the eruption. We also applied Sentinel-1 SAR data to look for ground deformation patterns on the volcano and the surrounding areas. The SO2 flux time series shows a slight increase in the degassing the days before the eruption; however, the variability was within the bounds of that seen during the previous months, showing no clear eruption precursors. No deformation is visible in the pre-eruptive interferograms. However, syn-eruptive deformation is visible in the first interferogram post-eruption. The deformation patterns suggest the emplacement of a dike on the area of Goma. Since the eruption began without any clear precursor from either data set, the risk posed by the volcano from another short notice eruption is high. The dike emplacement below the city increases the chances of a new eruption leading to more people being displaced from their houses to prevent further fatalities.

In recent years, many interests grew about the emergency management concerning natural events, such as volcanic eruptions, in parallel with the development of ever more refine satellite sensor technologies and machine learning algorithms. To address the challenging demands, which such events involve, fast and reliable procedures are needed. Neural Networks (NNs) have the potential to represent complex situations, as those related to natural phenomena, due to their capability to describe non-linear relationship between variables. Furthermore, they are characterized by a short processing time, a necessary condition for the analysis of these events in near real time. The reliability of NN-based algorithms basically depends on the quantity and the quality of the training dataset from which the net has to learn. Satellite instruments, with their large coverage and capacity to collect data from afar, are the necessary devices to catch the spatially extensive and occasional event as volcanic eruptions in particular in the most remote regions of the world. In our work we address the volcanic ash detection in near real time task by means of neural networks trained with MODIS (MODerate resolution Imaging Spectroradiometer) data and applied to Sentinel-3/SLSTR (Sea and Land Surface Temperature Radiometer) measurements. As test case the 2019 Raikoke eruption has been considered. The volcanic cloud detection procedure consists in identifying the ashy pixels in an SLSTR satellite image. The ash detection can be useful to face the above-mentioned issue related to the emergency phase of an eruption event, in which no-fly zone should be determined as soon as possible in order to guarantee the aviation security. Ash particles may indeed cause engine stalls and other damages to other part of the aircraft. The overall workflow of the detection procedure we developed is described hereafter and consists basically in generating the dataset, training the NN and finally applying the model. MODIS data from the 2010 Eyjafjallajökull (Iceland) eruption have been used to extract the training examples by applying a semi-automatic procedure, which exploits MODIS radiances and standard products. The patterns thus created have been used to train the model. Then the NN trained has been used to classify SLSTR data from the 2019 Raikoke (Russia) eruption. The output of the model is an image fully classified in eight classes: ash over sea, over land and over clouds, sea, land and ice surfaces, weather ice and water vapour clouds. The results of the classification are consistent and promising, given the challenging scenario with a wide distribution of meteorological clouds and considering the spatiotemporal geographical differences between data used for training the neural network and those on which the model was applied. However, further developments are under consideration in order to improve the NN accuracy and ability to generalize even over other eruptive scenarios. The results presented in this work have been obtained in the sphere of the VISTA (Volcanic monItoring using SenTinel sensors by an integrated Approach) project, funded by ESA and developed within the EO Science for Society framework [https://eo4society.esa.int/projects/vista/].

With a nearly continuously effusive eruption since 1983, the Kilauea volcano (Hawaii, USA) is one of the most active volcanoes in the world. At the beginning of May 2018, a sequence of eruptions on the Lower East Rift Zone (LERZ) caused an enhanced outbreak of volcanic gases and aerosols, releasing them into the troposphere. Since these gases and particles affect climate, environment, traffic, and health on regional to global scales, a continuos monitoring of the emission rates is essential. As satellites provide the opportunity to observe and quantify the emissions remotely from space, their contribution to the monitoring of volcanoes is significant. The TROPOspheric Monitoring Instrument (TROPOMI) onboard the Sentinel-5 Precursor satellite was successfully launched by the end of 2017 and provides measurements with unprecedented level of details with a resolution of 3.5 x 7 km2. This also allows for an accurate retrieval of trace gas species such as volcanic SO2. Here, it will be shown that the location and strength of SO2 emissions from Kilauea can be determined by the divergence of the temporal mean SO2 flux. This approach, which is based on the continuity equation, has been demonstrated to work for NOX emissions of individual power plants (Beirle et al., Sci. Adv., 2019). The present state of our work indicates that emission maps of SO2 can be derived by the combination of satellite measurements and wind fields on high spatial resolution. As the divergence is highly sensitive on point sources like the erupting fissures in the 2018 Kilauea eruption, they can be localized very precisely. The obtained emission rates are slightly lower than the ones reported from ground-based measurements in other studies like the one from Kern et al. (Bull. Volcanol., 2020). The effects of suboptimal conditions like high cloud fractions on the method probably affect the derived emission rates and have to be further analyzed. For this reason, aerosol and cloud information from the Visible Infrared Imaging Radiometer Suite (VIIRS) onboard the Suomi National Polar-orbiting Partnership (NPP) satellite are currently evaluated.

We have developed a new toolkit which allows the height and age of volcanic gas emissions to be quantified for each pixel containing a volcanic plume in a tropomi so2 image. This toolkit use meteorological wind fields and the location of the volcano generating the plume to determine the height and age of plume gas in each pixel. In this paper we review the current and future applications of this methodology. PlumeTraj allows us to quantify the temporal evolution of explosive eruptions, whose injection height reflects the mass eruption rate. The combination of mass eruption rate and gas emission rate allows us to quantify the gas content within the magma which powers the eruption, a key parameter to understand volcanic processes. The temporal pattern of mass eruption and gas emission rates is controlled by magma reservoir and magma ascent processes, allowing insight into the subsurface and correlation with geophysical observations of ground deformation with InSAR. Future work will incorporate retrieved heights from spectroscopic analysis of tropomi data, which will provide further constraints on the temporal evolution of volcanic gas emissions. This can then be applied in near real-time to provide volcanic risk managers with critically important information. PlumeTraj also provides high quality source terms for the location and quantity of gas, where height determination is important to accurately quantify SO2 mass. This allows better forecasts of the future dispersion of the gas. We can apply the same approach to images of volcanic ash retrieved from sensors such as IASI, providing high quality ash concentration forecasts. Volcanoes release large quantities of gas in the absence of explosive activity, and variations in emission rates often precede eruptions. PlumeTraj allows strongly degassing volcanoes flux to be quantified daily/hourly, complementing ground based monitoring. This can be integrated into the routine activity of volcano observatories. Averaging of tropomi images reveals weaker plumes, and allows many more volcanoes to be monitored and anomalous precursors degassing to be detected. We are therefore at the start of a new paradigm in the application of satellite data in volcanology research and monitoring which will provide powerful new scientific and risk management capacities. The planned future developments in earth observation satellites, particularly those in geostationary orbits, offers still further improvements in temporal resolution. The major challenges we face are data processing and management of the huge data streams produced by plumetraj. This requires a data science driven approach to realise the full global potential of the methodology.

Since the Eyjafjallajökull 2010 eruption and its massive disruption to the normal daily activity of the Northern Hemisphere, the satellite remote sensing community has focused their efforts for determining the total ejected SO2 mass and to forecast its possible movement and height. Space-born volcanic SO2 observations date back to the Cerro Azul, Galápagos Islands, 1979 eruption, sensed by TOMS/Nimbus-7, and currently a relatively large number of operational satellite sensors observe both volcanic, and also anthropogenic, SO2 emissions. The dependable estimation of the actual SO2 plume height, absolutely necessary for aviation safety and public health advisory, is still work-in-progress. Retrieval algorithms developed so far performed direct fitting and optimal estimation techniques to determine the height information. These are computationally very expensive and cannot not be applied in near-real time operational retrievals, in view of current and future satellite UV instruments with high resolution and related high data amount. In this work, we will present the inter-comparison and validation against independent datasets, of a combined principle component analysis and neural network retrieval algorithm called ‘Full-Physics Inverse Learning Machine’ [Hedelt et al., 2019.] The algorithm performs an extremely fast, ~3ms per S5P/TROPOMI pixel, yet accurate (

In response to the GEMS Validation Team AO call, this work discusses the first geophysical validation of GEMS Level-2 data products of NO2, HCHO, and O3 for four processed months (March to July 2021). The validation/evaluation methodology relies on the analysis of data retrieval diagnostics and on comparisons with correlative measurements acquired by the established ground based instrument types: MAX DOAS, Zenith Scattered Light DOAS, Pandora, Fourier transform infrared spectrometers (FTIR), Brewer, Dobson, and ozonesondes. Most of these instruments perform network operations in the framework of WMO's Global Atmosphere Watch (GAW) and its contributing networks NDACC, PGN and SHADOZ. Other validation data are obtained through direct contacts with instrument PIs, contributing on a best effort basis. A harmonized validation approach is intended for the different species. This will be achieved by building on the experience acquired during an uninterrupted series of satellite validation projects initiated in the early 1990s with the first UV visible DOAS nadir instrument in orbit, ERS-2 GOME, and continued with SCIAMACHY, OMI, the GOME-2 series and TROPOMI. The validation is carried out by the team who coordinates the operational validation of trace gas data from Sentinel-5 Precursor TROPOMI (within the S5P MPC) and Metop GOME 2 (within EUMETSAT’s AC SAF). In the last two decades, the team has developed methods and tools to assess and correct for multi-dimensional effects of smoothing and sampling in atmospheric composition remote sensing, and has identified validation challenges raised by the change in paradigm from LEO to GEO observations. In this perspective, this work on GEMS also serves as a pathfinder for the upcoming validation of the Copernicus Sentinel-4/5 constellation. After adaptation of the currently operational in-house satellite validation systems, a list of GEMS data Quality Indicators (QI) are derived from complementary investigations: (1) Data and information content studies based on the analysis of the GEMS data products, retrieval diagnostics and ancillary parameters; (2) Traceable preparation of the GEMS data and correlative measurements in view of data comparisons: co-location studies, unit and representation conversions, handling of smoothing and sampling issues, (sub)column integration; (3) Data comparisons leading to statistical estimates of the systematic bias and dispersion between GEMS and monitoring network data as a function of latitude or pollution regimes, their cycles, and their dependence on measurement and atmospheric parameters (e.g., clouds, solar zenith angle, aerosols, and slant column density); (4) Evaluation of the GEMS ex-ante uncertainties reported in the data files, with respect to the ex-post combined uncertainty estimates derived from the data comparisons, and their assessment with respect to the mission requirements.

Atmospheric gravity waves transport energy and momentum throughout the atmosphere and can travel large horizontal and vertical distances from the troposphere to the mesosphere and above. They contribute to atmospheric dynamics and among others drive the meridional pole-to-pole circulation in the mesosphere. Thus, knowing about the gravity waves, their spatio-temporal characteristics and their interaction with other waves and the atmospheric background is attracting more and more attention in order to further improve climate and even meteorological models. Gravity waves can be observed particularly well in the upper air layers of the atmosphere, especially in the upper mesosphere / lower thermosphere region. Measuring these waves especially in the mesosphere, however, is somewhat challenging. The OH airglow layer around the UMLT-height region offers good observation conditions and is often utilized for continuous observations at night times in this altitude region. The OH airglow layer is a chemiluminescent layer with strong emission in the short-wave infrared spectral range located in about 87km altitude with a layer halfwidth of about 4km. The OH airglow intensity is modulated by traversing atmospheric gravity waves which change the chemical reaction rates e.g. by temperature and pressure changes and by raising and declining the layer locally. Observing OH airglow with short-wave infrared imagers allows characterizing the gravity waves. In this study we present measurements of two ground-based FAIM (Fast Airglow IMager) systems, fast cameras sensitive in the short-wave infrared region observing the OH airglow layer. The cameras are located at Oberpfaffenhofen, Germany and Otlica, Slovenia, about 300km apart of each other and are pointing to the same volume in 87km altitude located in the Alpine Region above Northern Italy. Normally it is only possible to get horizontal wave characteristics (horizontal wavelength and horizontal propagation direction, wave period, phase speed) from single instrument OH airglow image sequences. We developed a novel tomographic algorithm to allow for a three-dimensional reconstruction of the airglow layer by combining images from the two viewing angles. In order to solve the highly underdetermined equation system, prior knowledge of the OH airglow layer vertical profile from multi-year observations of SABER on the TIMED-satellite are used on a statistical basis. This allows us, among others, to derive the vertical wavelength of the waves and see their three-dimensional structure. From that knowledge, further wave parameters can be estimated via the wave’s dispersion relation including the horizontal wind along the wave propagation. We will explain the tomographic reconstruction method which might in future also be applied on OH airglow image observations taken from satellite constellations. We will present a detailed case study and statistics about the wave parameters. This work received funding from the Bavarian State Ministry of the Environment and Consumer Protection.

The ESA atmospheric Validation Data Centre (EVDC) is the official ESA repository for calibration and validation (Cal/Val) data, it provides an online information system supporting users to exploit campaign datasets for Earth Observation missions and applications in the atmospheric domain. The EVDC portal (https://evdc.esa.int/) offers several tools supporting the user in terms of Cal/Val data query, data upload/download, format conversion (GEOMS conversion routines) and for production of ECMWF parameter’s maps. The EVDC platform also provides an access to satellite data for specific missions, in particular the system supports new atmospheric composition/dynamic missions namely Sentinel-5P, Aeolus and, in the near future, EarthCARE. The portal can be easily expanded to support new campaigns and satellite missions. Data exchange with the EVDC is regulated by a protocol with the aim to ensure data ownership, to prevent re-distribution to third parties and to protect intellectual properties. EVDC provides, moreover, a service to issue DOIs and Landing Pages for Cal/Val datasets. Following the feedback from the users, the Cal/Val data filtering capabilities and search performance have been enhanced. Users are also now able to preview and inspect the contents of the Cal/Val files before downloading. The new ECMWF forecast data extraction and plotting tool is designed to simplify the access to the ECMWF forecasts. The searchable satellite data archive includes the initial set of AEOLUS data products and the EVDC team continue to support researchers working on AEOLUS Cal/Val projects. As part of continued effort to make the EVDC platform more accessible we are now providing video tutorials (see the last EVDC Newsletter: https://evdc.esa.int/documentation/newsletters/) explaining how to work with most of the tools made available through the EVDC Web Platform. The EVDC team provides support to all currently registered users as well as new users interested in the data and functionalities offered by the EVDC portal and tools. Users are encouraged to contact the team via Support tab on the left-hand side in the EVDC portal with any questions, issues, bug reports and suggestions of improvements. One of the main goals of the EVDC project is to continue the development of the tools and functionalities according to the feedback from users.

In order to ensure that products delivered by air quality satellite sensors meet user requirements in terms of accuracy, precision and fitness for purpose, it is essential to develop a robust validation strategy relying on well-established and traceable reference measurements. In this context, the ESA Fiducial Reference Measurements for Ground-Based DOAS Air-Quality Observations (FRM4DOAS) activity is aiming at further harmonizing Multi-Axis Differential Optical Absorption Spectroscopy (MAX-DOAS) measurements. Since it provides vertically-resolved information on atmospheric gases at an horizontal scale approaching the one from nadir backscatter satellite sensors, the ground-based MAX-DOAS technique has been recognized as a valuable source of correlative data for validating space-borne observations of atmospheric species such as NO2, HCHO, SO2, O3, etc. Here we present the status of the first near-real-time (24h latency) central processing service for MAX-DOAS instruments that has been developed in the framework of the FRM4DOAS activity and which is operated as part of the Network for the Detection of Atmospheric Composition Change (NDACC). Since November 2020, the processing system, which includes state-of-the-art retrieval algorithms, delivers on a daily basis tropospheric NO2 vertical profile and total O3 column data from about 15 stations to the NDACC Rapid Delivery and ESA Validation Data Centre (EVDC) databases. The main aspects of the service development, like the central processing algorithm optimisation, operationalisation, and validation, data QA/QC, data policy, and the main lessons learned during the first months of operation, will be discussed. The status of the other FRM4DOAS products (tropospheric HCHO and stratospheric NO2 vertical profiles) that need to be further consolidated, as well as the new products that will be developed as part of the FRM4DOAS-2.0 R&D follow-up project, will be also presented. The NDACC MAX-DOAS central processing service and its future upscaling in terms of stations and data products will ensure that MAX-DOAS observations at a FRM quality level will be made available for the validation of present and future satellite missions like the Copernicus atmospheric Sentinels (5p, 4, 5).

The Boundary-layer Air Quality-analysis Using Network of Instruments (BAQUNIN) supersite is presented. The site has been collecting pollutant concentrations and meteorological parameters since 2017. Currently, BAQUNIN consists of three sites located in the city centre of Rome (Italy), and in the neighbouring semi-rural and rural areas. To the best of our knowledge, BAQUNIN is one of the first observatories in the world to involve several passive and active ground-based instruments installed in multiple measuring locations, managed by different research institutions, in a highly polluted megacity not far from the Tyrrhenian coast. BAQUNIN has been promoted by the European Space Agency to establish an experimental research infrastructure for the validation of present and future satellite atmospheric products and the in-depth investigation of the planetary and urban boundary layers. In this contribution, the main characteristics of the three sites are described, providing information about the complex instrumental suite and the produced dataset. Direct access to data and documentation is open to the citizen and scientific community at https://www.baqunin.eu. Specific datasets are available through EVDC (https://evdc.esa.int/), EUBRWENET (http://www.eubrewnet.org/eubrewnet), PGN (https://www.pandonia-global-network.org/) and EUROSKYRAD (http://www.euroskyrad.net/) international networks.

Atmospheric Research Infrastructures (RI) constitute an important resource of atmospheric data and related services. Parts of it have already been used for, e.g., satellite validation, but the full potential of the RIs is far from being exploited so far. The main advantage of an RI compared to project consortia is its coordination, long-term operation and sustainability. The associated observation facilities are selected and regularly monitored by the RI, standard protocols are used, and all data products are quality controlled in a traceable and harmonized way. This makes the activity highly predictable, and the data comparable over long time periods and large territories. Moreover, RIs enable science beyond their perimeter and put in place programs for granting wider user communities' access to data, facilities and services.     International organizations such as the European Space Agency (ESA), Copernicus, European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT), or European Environment Agency (EEA), generate and utilise vast amounts of data on atmospheric composition. While there is no doubt that they could benefit from access to Atmospheric RIs, to date, they have rarely approached their users to access the facilities through Trans-National Access (TNA) programs. This paper presents current opportunities offered through the joint Trans-National Access program of ACTRIS (Aerosol, Clouds and Trace gases Research InfraStructure), ICOS (Integrated Carbon Observation System) and IAGOS (In-service Aircraft for a Global Observing System) in the framework of the EC-funded ATMO-ACCESS project. The general approach is to explore various options for making access more attractive and efficient for international stakeholders, by combining the services of the individual RIs and jointly exploiting the facilities operated by observation platforms, exploratory platforms and central laboratories. The project aims to collect specific technical and scientific needs, to identify and/or develop research and technology services in response to these needs, to propose new flexible access modalities with novel procedures for access and service provision, which are more suited to the international organizations and their users (e.g., recurrent access, sequential access, multi-facility access, fast-track access, etc.), and to propose and design tailored experiments to be implemented as benchmark access pilots. Key aspects that have been identified include: tailored measurement protocols, mobilized       reference instruments for short-term campaigns, calibration services, developing and testing of new techniques, modules and algorithms, and addressing new synergies. In practice, 43 high-class observational and exploratory facilities operated by ACTRIS, ICOS and IAGOS in Europe (and beyond) will be open and provide their services for experiments and campaigns. We welcome ESA, Copernicus, EUMETSAT, EEA and other relevant organizations and scientists worldwide to actively contribute in designing the access pilots and giving recommendations for their implementation, with the aim to improve the TNA framework for the future benefit of atmospheric science.

Several operational validation systems for EO data are in use today, for instance the VDAF Automated Validation Server (VDAF-AVS) run in the ESA/Copernicus Sentinel-5p Mission Performance Centre (S5P MPC), and BIRA-IASB’s Multi-TASTE system for trace gases applied in various ESA activities and in EUMETSAT’s AC SAF. These follow generic validation protocols that rely on the synergistic use of ground-based, other satellite, and model data. Despite the good maturity of validation methods achieved for a few atmospheric species, methodological advances are still needed to arrive at true metrological traceability and inter-operability of the quality information [1,2,3]. In particular, the assessment of the ex-ante uncertainties provided with the satellite data products remains relatively uncharted territory. In this contribution, we identify key challenges that complicate the validation of ex-ante uncertainties, and demonstrate how we can progress on that with applications of advanced methods to Sentinel-5p TROPOMI data. These include: the use of a chi-squared statistical compliance test to quantify the satellite-to-ground agreement in terms of their reported ex-ante uncertainties, the use of model-based OSSE simulations with multi-dimensional metrology – with illustrations with the BIRA-IASB OSSSMOSE system [4,5] – to quantify spatiotemporal co-location mismatch and spatiotemporal sampling uncertainties in level-3 data sets, and the use of the triple co-location method to assess uncertainties even in the absence of ex-ante estimates [6]. Besides demonstrating the added quality information that can be obtained with these advanced methods, we also discuss the gaps that are revealed in the ex-ante uncertainty characterization of both satellite and ground-based measurements, in particular regarding the differentiation between uncertainties associated with systematic and random effects, and how these are now being addressed for instance in EuBrewNet and the Pandonia Global Network following the latest metrological concepts. Acknowledgements This work has been supported by the EC H2020 GAIA-CLIM (grant 640276) and Copernicus Cal/Val Solution (CCVS, grant 101004242) projects, and by the ESA/Copernicus Sentinel-5p Mission Performance Centre (S5P MPC). References [1] Loew, A., et al.: Validation practices for satellite-based Earth observation data across communities, Rev. Geophys., 55, 779– 817, doi:10.1002/2017RG000562, (2017). [2] H2020 GAIA-CLIM Gap Assessment and Impacts Document (GAID), http://www.gaia-clim.eu/page/gaid [3] H2020 CCVS: Towards a Copernicus CAL/VAL Solution, https://ccvs.eu/ [4] Verhoelst, T., et al. : Metrology of ground-based satellite validation: co-location mismatch and smoothing issues of total ozone comparisons, Atmos. Meas. Tech., 8, 5039–5062, https://doi.org/10.5194/amt-8-5039-2015, (2015) [5] Lambert, J.-C., et al.: “OSSSMOSE: Multi-purpose Simulator with Explicit Metrology for Global Observing Systems of Atmospheric Composition”, in the 42nd COSPAR Scientific Assembly, (2018). [6] Hubert, et al.: TROPOMI tropospheric ozone column data: Geophysical assessment and comparison to ozonesondes, GOME-2B and OMI, Atmos. Meas. Tech. Discuss. [preprint], https://doi.org/10.5194/amt-2020-123, in review (2020).

Pollutant gases information can be retrieved from ground-based and satellite visible and UV spectra, exploiting the Differential Optical Absorption Spectroscopy (DOAS) technique. The validation of the satellite NO2 tropospheric column shows the particular importance that should be given to measurements acquired in polluted regions. As the Po Valley (Italy) is one of these regions, it often experiences very high NO2 concentrations due to industrial, agriculture and urban activities and to the fact that the surrounding mountains protect the valley from winds that might otherwise disperse such pollution plumes. Despite this, an instrument compliant with the Fiducial Reference Measurements for Ground-Based DOAS (FRM4DOAS) standards is not yet present in Po Valley. Hence, the purpose of the DOAS-BO WPs within the QA4EO ESA project, that we will present, is to make a step to close this gap. We started using an existing custom spectrometer, named TROPOGAS and located in Bologna, that exploits the heritage of CNR-ISAC in spectrometers development. It operated with a measurement configuration that is as much as possible compliant with FRM4DOAS requirements. We assessed the performances of the TROPOGAS spectrometer during a measurement campaign held in Bologna at CNR-ISAC premises. The campaign focused on the evaluation of the synergies between ground-based remote sensing, in-situ near-surface (NOx Chemilunescence analyzer) and satellite data (TROPOMI). A second campaign, to inter calibrate the TROPOGAS with other ground-based instruments (i.e., PANDORSA 2S spectrometer, Brewer spectrophotometer), was foreseen within the Boundary-layer Air Quality-analysis Using Network of Instrument (BAQUNIN) supersite located in Rome. Recently the CNR-ISAC acquired two SkySpec2D spectrometers. One will be located at the CNR-ISAC Rome Atmospheric obServatory (CIRAS) and the other one at the San Pietro Capofiume (SPC) “Giorgio Fea” station. This type of instrument is fully compliant with FRM4DOAS requirements. For this reason, after an inter comparison with the TROPOGAS, the SkySpec2D of SPC will be used for the BAQUNIN campaign. The objectives of the QA4EO DOAS-BO WPs are to demonstrate the importance of the DOAS measurements in the Po Valley, re-enforce the Italian know-how on Multi-Axis (MAX)-DOAS technique and go towards the provision of standardized data for validation networks.

A new blackbody, named Hemispherical Blackbody (HSBB), has been developed to allow accurate calibration of broadband infrared detectors having a hemispherical acceptance angle. The aim of the HSBB is to significantly reduce the uncertainty of longwave downward radiation measurements traceable to the SI. Ground-based measurements of the longwave downward radiation are performed on weather stations globally and typically take place with so-called pyrgeometers having a hemispherical acceptance angle. Such measurements are, for example, organised within the Baseline Surface Radiation Network (BSRN) [1]. The Tilted Bottom Cavity Blackbody BB2007 [2] at Physikalisch-Meteorologisches Observatorium Davos / World Radiation Center (PMOD/WRC) has long provided the reference for tracing longwave downward radiation measurements within the BSRN to the SI. In recent years and within the EU co-founded MetEOC [3] series of projects, the HSBB has been developed with the objective to reduce the uncertainty of longwave downward radiation measurements. In the development and realisation of the HSBB, special attention was given to the requirements for calibrating pyrgeometers and similar detectors such as IRIS [4] instruments. To find an optimal design candidate for the cavity geometry and coating, effective emissivity simulations were carried out. The aim is to realise a blackbody cavity having the same effective emissivity for normal incidence and for the hemispherical opening angle in order to achieve consistency between measurements with a radiation thermometer and detectors with a hemispherical opening angle. Based on the outcome of the simulations and further considerations, it was decided that the blackbody will consist of a black coated cone in combination with a highly specularly reflecting golden hemisphere. Being directly traceable to the radiation temperature scale of the Physikalisch-Technische Bundesanstalt (PTB) [5] and therefore to the ITS90 and the SI, a second independent path of traceability for the BB2007 could be established. This was achieved by performing blackbody comparison measurements between the BB2007 and the HSBB. Those measurements took place as a measurement campaign at PMOD/WRC in autumn 2020. The new traceability path will improve and validate the existing traceability of the BB2007 which is based on contact thermometry and optical simulations. The development and main technical features of the HSBB will be demonstrated. Even so, the focus of the presentation will be on the blackbody comparison measurements and the objective of improving and validating the existing traceability path of longwave downward radiation measurements. References [1] The “Baseline Surface Radiation Network” (BSRN). https://bsrn.awi.de. Alfred Wegener Institute, Helmholtz Centre for Polar and Marine Research. [2] J. Gröbner. “Operation and Investigation of a Tilted Bottom Cavity for Pyrgeometer Characterizations”. In: Applied Optics 47.24 (2008), pp. 4441-4447. [3] Metrology for Earth Observation and Climate. www.meteoc.org. [4] J. Gröbner. “A Transfer Standard Radiometer for Atmospheric Longwave Irradiance Measurements”. In: Metrologia 49.2 (2012), pp. S105-S111. [5] I. Müller, A. Adibekyan, K. Anhalt, C. Baltruschat, B. Gutschwager, S. König, E. Kononogova, C. Monte, M. Reiniger, S. Schiller, D. R. Taubert, D. Urban and J. Hollandt. „Non-contact temperature measurement at the Physikalisch-Technische Bundesanstalt (PTB)”. In: Quantitative InfraRed Thermography Journal 18.3 (2021), pp. 187-212.

Current anthropogenic emission inventories generally rely on “bottom-up” approaches which are often highly uncertain due to their low temporal and spatial resolution since geographical and statistical data of different sources are mainly used to compile them infrequently. As a result, these inventories are not suitable for studies that refer to recent years, since are not available in near-real-time. On the other hand, a “top-down” approach takes advantage of space-borne atmospheric observations in order to infer emissions. This method is gaining more and more ground since satellite observations with high spatial and temporal resolution have become available. In this study, S5P/TROPOMI NO2 observations are employed in order to estimate updated NOx emissions over four of the largest power plants in NW Greece. NOx emissions in the region have been reported to significantly decrease in recent years due to the country’s targets set by the Directive 2016/2284/EC concerning the reduction of certain national atmospheric pollutants. LOTOS-EUROS chemical model simulations were used in order to find the best matching emissions to the S5P/TROPOMI observations through a data assimilation technique. The findings are validated by comparing NO2 surface simulations from both free and assimilated CTM model runs to NO2 in situ measurements from an air quality station near the largest power plant. The assimilated simulations suggest a strong decrease in NO2 concentrations compared to the free run and a much lower bias of ~2μg/m3 compared to ~10μg/m3 respectively. The emissions inferred also show an expected decrease of ~-60% supported by the reported decreases in the power plant energy production (~-50%) for the same period. The findings of this study suggest that S5P/TROPOMI high resolution NO2 observations can be used in a validated CTM in order to improve emission trends of local large emitters.

Lockdown in Pakistan due to global pandemic disease COVID-19 have reduced air pollution in mega-cities (Lahore, Karachi, Islamabad and Peshawar) during March, April and May, 2020. A prominent decrease in Nitrogen dioxide (NO2) concentrations was observed over mega-cities during the period March-May of 2020 in comparison with the same months of the previous year 2019. Satellite observations from Sentinel-5P over these cities indicate that the NO2 dramatically reduced in March 2020 as compared to the same month of previous year, with a maximum decrease of ~16% in Lahore. In April 2020, a remarkable decline in NO2 concentrations was observed over Lahore from 102×10-6 mol/m2 to 87×10-6 mol/m2, Karachi from 101×10-6 mol/m2 to 91×10-6 mol/m2, Peshawar from 76×10-6 mol/m2 to 75×10-6 mol/m2 and Islamabad from 93×10-6 mol/m2 to 82×10-6 mol/m2 as compared to the previous months of 2020 caused by diminished transport, industrial and economical activities during the lockdown. Satellite observations over mega-cities also showed that Aerosol Optical Depth (AOD) dramatically reduced in 2020 during the lockdown as compared with corresponding months of 2019, with a maximum decrease of ~52% in Lahore. Ground-based monitoring networks show a marked decrease in Particulate matter (PM2.5) concentration of ~40.4% in Lahore, ~48.6% in Karachi, ~22.5% in Islamabad and ~37.1% in Peshawar due to mandatory restrictions during lockdown as compared with pre-lockdown.

In this research we apply a new lane separation methodology for the maritime sector emissions attributed to the different vessel types and marine traffic loads in the Mediterranean and the Black Sea defined via the European Marine and Observation Data network (EMODnet). This methodology is implemented for the first time on the Sentinel-5 Precursor TROPOspheric Monitoring Instrument (TROPOMI) nitrogen dioxide (NO2) tropospheric vertical column densities, on the open source LOTOS-EUROS (Long Term Ozone Simulation – EURopean Operational Smog) chemical transport model simulations and on the Copernicus Atmospheric Monitoring Service Global Shipping (CAMS GLOB SHIP v2.1) nitrogen dioxide (NOX) emissions inventory. Due to its high spatial resolution and improved signal-to-ratio, TROPOMI can detect the significant shipping lanes in the Mediterranean on both seasonal and annual temporal scales. Furthermore, with the implementation of this new lane separation methodology we can differentiate the tropospheric TROPOMI NO2 vertical columns originating from diverse shipping activities. More specifically, cargo vessels are the highest emitting sector in the Mediterranean and the Black Sea regions followed by fishing, tanker, passenger, and other vessels (pleasure crafts, sailing, law, etc.). These results are in high agreement with the model simulations after the implementation of the new lane separation methodology, with the spatial correlation of the annual maritime NO2 loads ranging between 0.93 and 0.98. On a seasonal basis, both observations and simulations report a common variability with the winter-time comparisons in excellent agreement for the highest emitting sector, cargo vessels, with the observations reporting a mean load of 0.98 ± 0.54 and the simulations of 0.81 ± 0.45x1015 molecules cm-2 and correlation of 0.8. In summertime, the simulations report a higher decrease in modelled tropospheric columns than the observations, however still with high correlation between 0.85 and 0.94 per sector. Finally, CAMS GLOB SHIP v2.1 NOX emissions show a similar distribution with the TROPOMI observations - cargo and tanker combine for approximately 80% of the total emissions followed by fishing, passenger, and other vessels – with the only difference being that the fishing vessels emissions accounting for the third highest emitting sector. This may be possibly attributed to the outflow of inland emissions detected by TROPOMI and are straightforwardly enhancing the shipping emissions which are mainly located offshore. These encouraging findings will permit us to proceed with creating a top-down emissions inventory for NOX shipping emissions using S5P/TROPOMI satellite observations and a data assimilation technique based on the LOTOS-EUROS chemical transport model.

Nitrogen dioxide (NO2) is an important pollutant gas that affects air quality and climate through its impacts on the formation of tropospheric ozone and secondary organic aerosols. Its primary source in the atmosphere is fossil fuel combustion, and therefore NO2 levels peak in regions with important anthropogenic activities. Satellite maps revealed the city of Antwerp in Flanders, Belgium, as one of the strongest NO2 hotspots worldwide. In this work, we investigate the NO2 distribution in Flanders, as recorded by ground-based concentration measurements, airborne data and remotely-sensed columns, with the help of the high-resolution WRF-Chem chemical-transport model as cross-comparison tool. The model is configured to allow two nested domains with spatial resolution changing from 5 to 1km, so as to pinpoint the fine-scale distribution of local sources, and up-to-date high resolution emission inventories for Flanders are implemented. The physical parameterizations in the WRF model are evaluated through comparison against ground-based data from two meteorological stations in the Antwerp region. The simulated NO2 distributions are compared against (1) hourly measured concentration values from monitoring stations in Flanders, (2) vertical columns measured by an airborne hyperspectral imager APEX, providing a 2-dimensional spatial mapping, on 27 and 29 June 2019, and (3) spaceborne NO2 columns obtained from the high-resolution TROPOMI sensor aboard S5p. The consistency of the model biases across the three datasets will be discussed. Recommendations will be made for improving model performance in this region. The TROPOMI NO2 biases against APEX data will be evaluated with the help of the model. Finally, the WRF-Chem model performance against TROPOMI and ground-based data will be discussed and characterized for a period of 15 days in June 2019.

The Tropospheric Monitoring Instrument (TROPOMI), aboard the ESA’s Copernicus Sentinel-5 Precursor (S5P) satellite, offers improved retrievals by combining the strengths of SCanning Imaging Absorption SpectroMeter for Atmospheric CHartographY (SCIAMACHY) and Ozone Monitoring Instrument (OMI) and by using state-of-the-art technology. In this study, tropospheric NO2 retrievals (Level-2) from TROPOMI have been used to observe current state of NO2 pollution and detect new hotspots in South Asia during July, 2018 to September, 2021. Also, a comparison of spatial distributions of NO2 retrieved from TROPOMI and OMI onboard Aura satellite has been made. The overall average values of the NO2 concentrations are 6.67x10-5 moles/m2 and 3.27x1015 molecules/cm2 for TROPOMI and OMI sensors respectively for the study period. TROPOMI-NO2 has the lowest average value in Jan 6.0x10-05±3.8x10-06 moles/m2 with 1.6% decreasing rate and the highest average value in Jun 7.6x10-05±4.5x10-06 moles/m2 with -5% decline rate. Over strongly polluted areas, TROPOMI retrievals show ability to effectively detect new and small size hotspots, and to segregate nearby mega sources. In Afghanistan, Kabul city is more clearly appearing as hotspot. In Pakistan, newly installed Sahiwal coal-based power plant was undetectable by OMI retrievals. Similarly, in India several new hotspots in eastern mining and power plants zone have been separately detected. However, in elevated eastern regions of Assam and Nagaland states (India), TROPOMI data shows insignificant concentrations and detects no hotspots of NO2 as compared to OMI.

We will present airborne measurements of biomass burning pollution trace gases in the upper troposphere from the Gimballed Limb Observer for Radiance Imaging of the Atmosphere (GLORIA). The origin of these polluted air masses is estimated with help of backward trajectories, and the vertically resolved GLORIA cross sections are applied to evaluate simulation results from the Copernicus Atmosphere Monitoring Service (CAMS) model. Simultaneous GLORIA observations of peroxyacetyl nitrate (PAN), ethane (C2H6), formic acid (HCOOH), methanol (CH3OH), and ethylene (C2H4) above the South Atlantic in September and October 2019 were obtained during the SouthTRAC (Transport and Composition in the Southern Hemisphere Upper Troposphere/Lower Stratosphere) campaign with the German High Altitude and Long range research Aircraft (HALO). On 8 September 2019, a filamentary structure with maximum volume mixing ratios (VMRs) of 900 pptv for PAN, 1100 pptv for C2H6, 800 pptv for HCOOH, 4000 pptv for CH3OH, and 200 pptv for C2H4 has been observed at altitudes between 7 km and 14 km. On 7 October 2019, one large plume has been measured at altitudes between 8 km and 13 km for all discussed species, except C2H4, besides smaller enhancements. Maximum VMRs of 1000 pptv for PAN, 1400 pptv for C2H6, 500 pptv for HCOOH, and 4500 pptv for CH3OH have been observed. With help of backward trajectories, we show that measured pollutants are likely originating from South America and central Africa. In the same regions, elevated PAN VMRs are visible at the surface layer of the CAMS model during the weeks before both measurements. In comparison to simulation results of the CAMS reanalysis data interpolated onto the GLORIA measurement geolocations, we show that the model is able to reproduce the overall structure of the measured pollution trace gas distributions. For PAN, the absolute VMRs agree with the GLORIA measurements, too. However, C2H6 and HCOOH are generally underestimated by the model, while CH3OH and C2H4, the trace gases with the shortest atmospheric lifetimes of the discussed pollution trace gases, are overestimated by CAMS. We assume the transport and locations of emission of the CAMS model to be appropriate, while emission strengths and atmospheric loss processes of all discussed trace gases except of PAN possibly should be improved for the model.

Increased fossil fuel combustion due to industrialization and traffic density has been the main source of nitrogen dioxide (NO2) in the atmosphere over Europe. The presence of NO2 in air leads to the formation of secondary pollutants such as tropospheric ozone and nitrate aerosols, negatively impacting human health and environmental conditions. Therefore, it is important to continuously monitor the distribution of NO2 concentrations over large regions to regulate the existing environmental policies for sustainable development. Remote sensing datasets from satellites such as Sentinel-5P (S5P) provide spatially continuous high resolution NO2 datasets. However, they typically provide only an integrated estimate of the tropospheric NO2 column and not of the surface concentration of NO2, which is the measure usually used for exposure and health applications. In this study, the objective is to derive a spatially continuous surface NO2 ­­concentration from tree-based machine learning models such as Random Forest and Extreme Gradient Boosting (XGBoost), using datasets such as tropospheric NO2 from the TROPOspheric Monitoring Instrument (TROPOMI) aboard S5P, day-night band from the Visible Infrared Imaging Radiometer Suite (VIIRS) instrument aboard the Suomi National Polar-orbiting Partnership (SUOMI-NPP) satellite, NO2 station measurements from the European Environmental Agency (EEA) and secondary variables such as meteorological parameters and land use land cover information. The preliminary outputs of this work show promising results with a mean absolute error of 10.4 µ/m3 and 12 µ/m3 respectively for Random Forest and XGBoost models. Furthermore, the feature importance ranking of input variables given by the tree-based machine learning models provide a base to effectively select only the relevant parameters for NO2 concentration estimation, thus reducing the data dimensionality and model complexity.

Carbon monoxide in high concentrations can have severe health effects. For this reason there are regulations for the monitoring of it, along with several other gases and particles with negative health effects. However, as the anthropogenic emissions of carbon monoxide have declined, the need for monitoring has diminished and no monitoring station in Finland measures it anymore.In this work we have used carbon monoxide measurements from TROPOMI instrument onboard Sentinel-5P. We have used the data to support air quality assessment in two different ways. First we calculated a long-term average over Finland to find the average CO distribution and to recognize areas with possible larger sources. For these areas we could assess the need for continuous monitoring. The long term average maps were also compared to emission database data.The other way we utilized satellite observations was to estimate ground level CO concentrations. We used the ground level data from scientific monitoring stations in Helsinki, Sodankylä and Pallas to calculate a simple linear relation between ground level concentrations and the total column concentrations provided by satellite measurements. Using this relation we evaluated yearly average and maximum ground level concentrations for specific monitoring regions. The average concentrations were lower for the northern Finland with lower population density and less traffic. However the maximum concentrations can be related to long range transport from for example an intense wild fire region and thus do not follow any clear pattern.

The first airborne in situ measurements of sulphur dioxide (SO2) emissions (plumes) from two coal-fired power plants in Bosnia-Herzegovina (Tuzla) and Serbia (Nikola Tesla) were carried out with the German research aircraft Falcon-20 in cooperation with local partners during the METHANE-To-Go field experiment in autumn 2020. Downwind of the power plants, SO2 mixing ratios exceeding 100 ppb were measured in a distance ~20-40 km from the sources. The plumes were trapped in well-defined inversion layers between ~500-1000 m altitude. Our airborne measurements can be used to validate synchronously, spaceborne SO2 measurements from the TROPOspheric Monitoring Instrument (TROPOMI) onboard the Sentinel-5P satellite. A first intercomparison indicates some problems with dense smoke clouds frequently covering these countries in winter. However, one part of the Nikola Tesla flight is well suited for TROPOMI-SO2 validation, since it was obtained during cloud-free conditions with a well-defined vertical extension of the probed SO2 plume (needed to estimate the Vertical Column Density, VCD). These airborne measurements and model simulations can also be used to determine the SO2 emission strength of the power plants. First estimates (mass balance approach) show that the SO2 mass flux from Tuzla is about twice as high as indicated by common emission inventories. Our outlook will give a first glance of further TROPOMI validation measurement attempts carried out with a Cessna Caravan aircraft in northern Scandinavia in August 2021 focusing on methane (CH4) from wetlands.

The simulation of Earth/atmosphere synthetic spectral radiances is crucial for the characterisation of clouds’ impact on weather and climate. It is, in fact, well recognized that cloud identification and their properties retrieval are essential elements in the description of the radiative balance within atmospheric layers. However, the solution of the full radiative transfer equation in a cloudy atmosphere is generally computationally expensive due to the complex modelling of multiple scattering effects. This study aims at investigating the level of accuracy of scaling methods and analytical approximations commonly used in radiative transfer codes to avoid the application of advanced and time consuming numerical solutions. This study is brought on within the initiatives supported by ESA and ASI for the preparatory studies of the ESA 9th Earth Explorer mission called FORUM (Far-infrared Outgoing Radiation Understanding and Monitoring), which will be set on orbit in 2026. FORUM will carry a Fourier Transform spectrometer capable to observe the spectrally resolved Earth’s emission spectra in the 100–1600 cm−1 band and, together with IASI-NG, it will cover the full long-wave spectrum of the Earth energy emission to space. Specifically, the present work focus on the Chou’s approximation, and a simple scaling method based on the similarity principle. In particular, the former one is widely implemented in existing fast radiative codes to solve the radiative transfer problem in the infrared spectral region. A widespread collection of atmospheric scenarios is considered. The top of the atmosphere synthetic spectral radiances are computed for each scenario considering alternatively an accurate and time consuming methodology, such as the discrete ordinate solution (DISORT) or the approximate methodologies. The residuals, calculated as the difference between the full scattering solution and the scaling methods, are evaluated at far- and mid infrared wavelengths and then compared with the goal noise of the FORUM satellite sensor. The results are discussed and analyzed in terms of geometrical, microphysical, and optical properties of the cloud layers. A range of validity within which these fast methodologies can be applied with sufficient accuracy is also provided. In particular, in case of both water and ice cloud scenarios, the approximate solutions are perform well in the mid infrared for most of the cases studied. When the far infrared region is considered, not negligible inaccuracies are observed. The implementation of innovative solutions for fast computations are also discussed.

The Far-Infrared Outgoing Radiation and Monitoring (FORUM) mission will provide an unprecedented opportunity to verify and potentially improve the ability of global circulation models (GCM) in simulating the Outgoing Longwave Radiation (OLR). Accurate simulation of the OLR is in fact crucial to better constrain the different radiative feedbacks. Planned for launch in 2026, FORUM will measure spectrally resolved radiances of the Earth’s emission spectrum at the top of the atmosphere (TOA) from 100 to 1600 cm-1 filling the existing observational gap of the far-infrared region, from 100 to 667 cm-1. In addition, FORUM will fly in loose formation with IASI-NG, which will continue to cover the middle infrared range of IASI from 645 to 2760 cm-1. In anticipation of FORUM measurements, we aim at comparing IASI existing observations to synthetic radiances extracted from the EC-Earth GCM (version 3.3.2), a recent European model based on ECMWF’s Integrated Forecasting System (IFS) for the atmosphere–land component and the ocean model NEMO, including sea ice (LIM2) and land surface components. Climate model outputs only provide simulated energy fluxes integrated over the whole Earth emission spectrum making difficult the detection of biases and the identification of possible compensation errors in the estimation of the OLR. Conversely, comparing simulated to observed spectra, allows to point out the potential model criticalities in specific spectral bands which contain the signatures of particular climate variables. In order to extract simulated spectra from the climate model, EC-Earth has been equipped with the Cloud Feedback Model Intercomparison Project (COSP), a simulator package able to map the model state into synthetic observations from different satellite-borne active (CloudSat (radar) and CALIPSO (lidar)) and passive (ISCCP, MISR and MODIS) sensors. We have further developed the package by implementing inside COSP the radiative transfer model σ-FORUM, a monochromatic code able to reproduce synthetic radiances in the Far-Infrared and Mid-Infrared regions compatible with future FORUM and existing IASI observations. Due to the high computation cost of the operation, the efficiency of the EC-Earth model equipped with the new COSP module has been improved by modifying the σ-FORUM original code structure and by reducing the original resolution of the model. Therefore, simulations provided by EC-Earth model equipped with the new COSP + σ-FORUM module have been performed with prescribed sea surface temperature and sea-ice cover every 6 hours, over a timeframe consistent with the availability of IASI METOP-A L1C data, from 2007 to 2016. The comparison between nadir radiances simulated by the EC-Earth climate model and the climatology built from ten years of IASI observations represents a very high confidence test for the direct verification and improvement of the GCM. The same approach could be extended to other climate models and, in the near future, it will involve FORUM measurements for a comprehensive analysis of the climate model ability in reproducing the whole Earth emission spectrum.

In 2019, the Italian Space Agency (ASI) started a new project entitled FORUM - scienza (FORUM science). One of the project objectives is to improve and share, within the interested National scientific community, the models needed for interpretation and analysis of the measurements of the forthcoming FORUM (Far-infrared Outgoing Radiation Understanding and Monitoring) experiment. FORUM will be the ninth Earth Explorer mission of the European Space Agency (ESA), scheduled for launch on a polar satellite in 2026. The core instrument will be a Fourier Transform Spectrometer covering the spectral region from 100 to 1600 cm-1 (from 100 to 6.25 microns in wavelength), with a full-width-half-maximum spectral resolution of 0.5 cm-1. The spectrometer will measure the Earth’s upwelling spectrum in the nadir-looking geometry. FORUM will fly in loose formation with the MetOp-SG-A satellite, hosting the Infrared Atmospheric Sounding Interferometer - New Generation (IASI-NG). The MIR (645 to 2760 cm-1) upwelling atmospheric spectrum measured by IASI-NG will complement the FORUM FIR spectrum, providing an unprecedented spectral coverage, from 100 to 2760 cm-1. While the FIR part of the spectrum (100-667 cm-1) measured by FORUM is the most sensitive to the water vapour content in the UTLS, the atmospheric windows in the MIR are measured by IASI-NG with a very high signal-to-noise ratio, thus supplying very precise information on the surface temperature and on the temperature profile, which are essential to constrain the retrieval parameters. Among the models improved in the frame of the FORUM - scienza project, σ-FORUM is a fast parametric radiative transfer code with the capability of simulating the upwelling spectrum, from 50 to 2810 cm-1, with sufficiently high spectral resolution (0.01 cm-1), in all sky conditions. Along with the spectral radiance, the model computes its Jacobians with respect to relevant atmospheric, cloud and surface parameters. The characteristics and the accuracy figures of the σ-FORUM model are presented by Masiello et al., within this conference. Based on σ-FORUM, we have built a fast and flexible Bayesian inversion model. This model includes the possibility of performing the synergistic inversion of both FORUM and IASI-NG measurements and to apply various types of constraints in the inversion, such as Optimal Estimation and / or Tikhonov regularization. Profile retrieval grids, together with the set of parameters to be retrieved are selected via input files. Currently, the following sets of parameters can be retrieved: temperature, water vapour and ozone profiles, surface temperature and surface spectral emissivity (at a selectable set of wavenumbers). The program also offers the possibility to minimize the effect of possible frequency scale errors of the spectrum by retrieving a frequency scale factor. The possibility to retrieve profiles of ice- and liquid- water contents and of droplet- and ice- effective dimensions will be implemented in the near future. In this work we illustrate the structure of the developed retrieval model, its performance, and the results of the first tests we made to optimize the retrieval setup in clear-sky conditions.

The forthcoming Earth Explorer 9 - FORUM (Far Infrared Outgoing Radiation Understanding and Monitoring) mission will provide global sounding of the Earth’s thermal emission to space in the Far-Infrared (FIR, from 100 to 600 cm-1). The on-board Fourier Transform Spectrometer (FTS) will deliver a wide set of TOA radiances during the mission lifetime that will represent a benchmark for future climate missions. Atmospheric water vapour plays a key role in modulating the OLR in the FIR hence the acquired spectra will also enable future improvements of FIR water vapour spectroscopy. The current uncertainty on spectroscopic data and continuum models is one of the main sources of error on the estimation of the atmospheric radiative properties in the FIR, especially in clear sky conditions where the strong absorption provided by the water vapour rotational band is almost totally responsible for the atmospheric extinction. In preparation for the FORUM-EE9 mission launch in 2026, the analysis of the consistency between observations from prototypes of the FORUM FTS with similar spectral resolution and noise performance and RTM simulations can be a first step toward an improvement of the water vapour spectroscopic data. Here we exploit the balloon-borne measurements above Teresina (Brazil) from the Radiation Explorer in the Far InfraRed – Prototype for Applications and Development (REFIR-PAD) and aircraft observations above the North Sea from the Tropospheric Airborne Fourier Transform Spectrometer (TAFTS). These measurements are used to assess the performances of different spectroscopic databases (HITRAN, AER, GEISA), coupled to different water vapor continuum parametrizations. The performances are assessed through a consistency analysis between measurements and simulations aimed to inform which best simulates the observed FIR spectra. The work will also demonstrate the capability of FORUM-like instruments to give a reliable contribution to spectroscopic improvements in the FIR.

FORUM (Far-infrared Outgoing Radiation Understanding and Monitoring) will fly as the 9th ESA's Earth Explorer mission, and an End-to-End Simulator (E2ES) has been developed as a support tool for the mission selection process and the subsequent development phases.The current status of the FORUM E2ES project is presented, together with the characterization of the capabilities of a full physics retrieval code applied to FORUM data. We show how the instrument characteristics and the observed scene conditions impact on the spectrum measured by the instrument, accounting for the main sources of error related to the entire acquisition process, and the consequences on the retrieval algorithm.Both homogeneous and heterogeneous case studies are simulated in clear and cloudy conditions, validating the E2ES against two independent codes: KLIMA (clear sky) and SACR (cloudy sky).The tests carried out show that the performance of the retrieval algorithm is compliant with the project requirements both in clear and cloudy conditions.The far-infrared part of the FORUM spectrum is shown to be sensitive to surface emissivity, in dry atmospheric conditions, and to cirrus clouds, resulting in improved performance of the retrieval algorithm in these conditions.The retrieval errors increase with increasing the scene heterogeneity, both in terms of surface characteristics and in terms of fractional cloud cover of the scene.

FORUM (Far-infrared Outgoing Radiation Understanding and Monitoring) experiment will be the ninth Earth Explorer mission of the European Space Agency (ESA), it is scheduled for launch on a Polar satellite in 2026. The FORUM core instrument will be a Fourier Transform Spectrometer covering both the Far-Infrared (FIR) and the Middle-InfraRed (MIR) spectral regions. It will measure, at nadir, the upwelling spectral radiance emitted by the Earth across the most relevant infrared part of the electromagnetic spectrum, from 100 to 1600 cm-1 (from 100 to 6.25 microns in wavelength), with a full-width-half-maximum spectral resolution of 0.5 cm-1. FORUM will fly in loose formation with the MetOp-SG-A satellite, hosting the Infrared Atmospheric Sounding Interferometer - New Generation (IASI-NG). The MIR (645 to 2760 cm-1) upwelling atmospheric spectrum measured by IASI-NG will nicely complement the FORUM FIR spectrum, providing an unprecedented spectral coverage, from 100 to 2760 cm-1. While the FIR part of the spectrum (100-667 cm-1) measured by FORUM is the most sensitive to the water vapour content in the UTLS, the atmospheric windows in the MIR are measured by IASI-NG with a very high signal-to-noise ratio, thus supplying very precise information on the surface temperature and on the temperature profile, that are essential to constrain atmospheric and surface parameter retrievals. In this study, we compare two alternative approaches to determine atmospheric and surface state parameters by exploiting simultaneously both FORUM and IASI-NG measurements. We examine the synergistic retrieval of state parameters from the simultaneous inversion of both measurements, and the a posteriori combination, via the Complete Data Fusion method, of the state parameters retrieved from the individual measurements. We perform test retrievals based on synthetic clear-sky measurements and characterize the quality of synergistic and fused products by means of their error evaluated both from the covariance matrix of the retrieval and from the statistics of the differences between retrieved and true state parameters. The retrieval products considered in this study include the profiles of water vapour and temperature, surface temperature and spectral emissivity from 100 to 2200 cm-1. As expected, we find that the exploitation of IASI-NG measurements permits to improve significantly the quality of FORUM retrieval products. These improvements, however, are progressively lost as soon as the mismatch between the two measurements increases. While synergistic retrieval and Complete Data Fusion provide almost identical results for perfectly matching measurements, differences comparable to the noise error are observed in the presence of time- and / or space- mismatches. We discuss the advantages and disadvantages of each of the two methods and analyze the limits of their applicability.

Generative Artificial Intelligence is a branch of Machine Learning algorithms. Generative Models such as Generative Adversarial Networks (GANs) and Autoencoders (AE) models can be used to generate data and manipulate images (e.g. Goodfellow et al., 2014). The generated data allow to produce large amount of data at reduced computing costs with respect to traditional Forward Models (FM) and to tests new configurations of variables. In the frame of the ESA DeepLIM project, GANs and AE are used to simulate Thermal Infra-Red (TIR) spectra acquired by satellite-borne instruments. The spectra are generated starting from a dataset of spectra simulated with a Radiative Transfer Model (RTM) or measured during an aircraft campaign. The simulated dataset consists on both Top Of the Atmosphere (TOA) and Bottom Of the Atmosphere (BOA) radiances. The generated dataset is composed by the same fields. To simulate the satellite measurements, both original and newly generated radiances are convolved with spectral response functions in the 5 TIR channel of the Sentinel Land Surface Temperature Monitoring (LSTM) mission (at 8.6, 8.9, 9.2, 10.9, 12.0 μm) and converted into Brightness Temperatures (BTs). Then, the extended dataset is processed to retrieve Land Surface Temperature (LST) in the five channels. The retrieval is performed using both “state-of-the-art” algorithms (e.g. TES algorithm (Gillespie et al., 1998)) and Neural Network (NN) We will present the characteristics (e.g. average values, standard deviations) of the generated dataset and compare them to the original ones. Then the results of LST retrieval will be presented to asses the usability of generated spectra for remote sensing applications. Although in DeepLIM project the generated spectra are used to retrieve surface parameters, the fact that the underling radiative properties of the original spectra are preserved suggest that the same features can be used to retrieve atmospheric variables. This highlights the capability of deep learning techniques to support missions and retrieval algorithm development. References Gillespie, A. R., S., Rokugawa, T., Matsunaga, J., S., Cothern, S., Hook, and A. B. Kahle (1998), A Temperature and Emissivity Separation algorithm for Advanced Spaceborne Thermal Emission and Reflection radiometer ASTER images. IEEE Transactions on Geoscience and Remote Sensing, 36, 1113–1126. Goodfellow,I., J., Pouget-Abadie, M., Mirza, B., Xu, D., Warde-Farley, S., Ozair, A., Courville, and Y. Bengio (2014), Generative adversarial nets, Advances in neural information processing systems, 2672–2680.

Satellite instruments can detect and quantify methane emissions from anthropogenic activities in remote parts of the world, where monitoring with ground-based observations is not feasible. These instruments include high spatial resolution multiband spectrometers capable of pinpointing individual point sources. The multiband instruments onboard Sentinel-2 and Landsat satellites can detect methane emissions from super-emitters as the resulting methane column enhancement at a spatial scale of 20 to 30 m near the source produces sufficient absorption signal integrated within the broader short-wave infrared bands. These bands are also sensitive to natural gas flaring, which can be used as an additional piece of information when methane leaks originate from unlit flares venting methane directly into the atmosphere. The observations acquired routinely from the multiband satellites have frequent revisit times and are suited for monitoring methane-leaking sources. Here we present a machine learning-based methane leak monitoring system that uses convolutional neural networks (CNN) to automatically detect methane plumes using Sentinel-2 data at numerous point sources across the globe.

The detection of methane point sources and the subsequent corrective actions, have been identified as an important climate change mitigation activity. In the last years, several studies have proven the capability of different satellite missions to map such methane plumes. Recently, Sentinel 2 (S2) has been shown to be useful for high resolution methane mapping. This study seeks to better understand the potential and limitations of the S2 mission for methane monitoring and is framed in the recent ESA HiResCH4 project intended to develop methods for high resolution methane mapping from space. The implemented retrieval methodology calculates methane concentration enhancements from pixel-level estimates of methane transmittance at S2 SWIR band 12 (2100-2300 nm). This band 12 is sensitive to the methane absorption and, in order to calculate methane transmittance, the band 12 spectral channel is normalised by a ‘methane-free’ reference band, i.e. with no excess methane present. Different methodologies are proposed to estimate a ‘methane-free’ reference band that include a band regression and/or temporal normalisation. The study is focused on the retrieval performance attending to factors such as the temporal misregistration, angular differences or spectral mismatch. Particularly, the temporal normalisation has been found to be effective not only to estimate the plume transmittance but also increases the revisit time by indistinctively setting Landsat and S2 mission overpasses as a 'methane-free' reference. In a second part of the study, the detection limit of the S2 mission is evaluated for a selected retrieval methodology at different scenes through realistic simulations of methane plumes. The detection limit is defined as a distribution of potential plume source rates per S2 scene that depend on the plume shape and the surface heterogeneity where it locates. Real data of known areas of methane emissions will be shown to verify the results obtained throught these simulations.

The fossil fuels that people are burning for energy cause a significant increase in atmospheric CO2 concentrations and the increase are responsible for about two-thirds of the total energy imbalance that is causing Earth's temperature to rise, which makes global accurate CO2 monitoring of great significance. The Greenhouse Gases Observing Satellite (GOSAT) is the world's first spacecraft to measure the concentrations of CO2 from space and it has high-precision hyperspectral atmospheric CO2 monitoring from 2009. Several atmospheric CO2 products, such as ACOS, NIES, OCFP and SRFP products from full physics retrieval algorithm, both provide CO2 surface flux of GOSAT. These products have different characteristics and advantages and have different performance in different regions. In order to obtain the CO2 data set with high precision, low uncertainty and high coverage, we used the products of the above four algorithms for fusion. The method of the paper directly uses the XCO2 variable of four products for fusion to retain the advantages of each algorithm. The XCO2 fusion process is carried out in each 1° × 1° grid of the world. Meanwhile, the influence of surface albedo and aerosol optical depth (AOD) which are two main factors in XCO2 retrieval, is comprehensively considered and quantified. The fusion algorithm proposed in this study consists of three parts: Firstly, in each partition grid, we choose the media XCO2 as the representative for each algorithm. Secondly, the uncertainty of each grid and each algorithm is determined according to AOD and surface albedo factors. Finally, the maximum likelihood estimation method is used to merge XCO2 products. Since the products of the same satellite are used, there is no need to consider the matching of time and space factors. The fusion results were encouraging, not only increasing the coverage of XCO2, but also demonstrating a good accuracy in validation with Total Carbon Column Observing Network observations. Index Terms: GOSAT, XCO2, maximum likelihood estimation, Fusion

Carbon dioxide (CO2) is the most important anthropogenic greenhouse gas and the main driver of global warming. Its atmospheric concentrations have risen more than 40% since pre-industrial times. Almost 90% of this increase results from fossil fuel combustion, emitting CO2 predominantly from localized sources. In order to control CO2 emissions it is necessary to accurately monitor them. Under the Paris Agreement, progress of emission reduction efforts is tracked on the basis of regular updates to national greenhouse gas (GHG) inventories, referred to as bottom-up estimates. Emission estimates can also be obtained top-down using atmospheric observations for verification and to obtain additional information. In this context, especially satellite observations are important as they can provide relevant information globally. Due to CO2's long lifetime and large fluxes of natural origin, the column-average concentrations resulting from anthropogenic emissions from individual source points are usually small compared to the background concentration, and these enhancements are often barely larger than the satellite's instrument noise. This makes the detection of CO2 emission plumes and the quantification of anthropogenic fluxes challenging. NO2 is co-emitted with CO2 in the combustion of fossil fuels. It has a much shorter lifetime, and as a result, its vertical column densities can exceed background values and sensor noise by orders of magnitude in emission plumes. This makes it a suitable tracer for recently emitted CO2. The objective of this study is to quantify the CO2 emissions from localized sources such as power plants by using XCO2 (the column-averaged dry air mole fraction of CO2) retrievals from the Orbiting Carbon Observatory 3 (OCO-3) in its snapshot area mode. Our presentation describes a semi-automatic procedure to select promising targets and overpasses, a plume detection method using NO2 as a tracer for recently emitted CO2 and an inversion technique to quantify CO2 emissions from detected CO2 plumes.

A large and fast view of volcanic plumes as detection and measurement of volatiles components exolving from craters are possible by using hyperspectral remote sensing if their absorption bands are in the sensor spectral range. In the present study the developed algorithm to calculate CO2 columnar abundance in tropospheric volcanic plume is presented. The algorithm is based on a modified CIBR 'Continuum Interpolated Band Ratio' remote sensing technique initially developed to calculate water vapor columnar abundance. The retrieval techniques exploit spectroscopy measurements by analysing gases absorptions features in the SWIR (Short Wave InfraRed) spectral range. Specifically, PRISMA (PRecursore IperSpettrale della Missione Applicativa) acquisitions are used for gases retrieval purposes. The PRISMA space mission was launched by the Italian Space Agency (ASI) on March 22, 2019; the on-board spectrometer is able to measure in the VNIR (0.4-1.0 µm) and SWIR (0.9-2.5 µm) spectral ranges, with a spatial resolution of 30 m. In this study, the inversion techniques is applied in order to compare its performances for CO2 detection and retrieval. The considered test sites are Stromboli and Etna volcanoes located in southern Italy and characterized by a persistent degassing and in the case of Stromboli by regular explosions from small to paroxysm.

Methane CH4 is an important anthropogenic greenhouse gas and its rising concentration in the atmosphere contributes significantly to global warming. Satellite measurements of the column-averaged dry-air mole fraction of atmospheric methane, denoted as XCH4, can be used to detect and quantify the emissions of methane sources. This is important since emissions from many methane sources have a high uncertainty and some emission sources are unknown. In addition, sufficiently accurate long-term satellite measurements provide information on emission trends and other characteristics of the sources, which can help to improve emission inventories and review policies to mitigate climate change. The Sentinel-5 Precursor (S5P) satellite with the TROPOspheric Monitoring Instrument (TROPOMI) onboard was launched in October 2017 into a sun-synchronous orbit with an equator crossing time of 13:30. TROPOMI measures reflected solar radiation in different wavelength bands to generate various data products and combines daily global coverage with high spatial resolution. TROPOMI's observations in the shortwave infrared (SWIR) spectral range yield methane with a horizontal resolution of typically 7x7km2. We use a 3-year XCH4 data set (2018-2020) generated with the WFM-DOAS retrieval algorithm, developed at the University of Bremen, to detect locally enhanced methane concentrations originating from emission sources.Our detection algorithm identifies temporally stable local enhancements relative to the surroundings by utilising different filter criteria, such as a certain threshold that the local methane anomalies must exceed.To attribute the detected methane enhancements to potential sources they are compared to inventories of anthropogenic methane emissions. In this presentation, the algorithm and initial results concerning the detection of local methane enhancements by spatially localized methane sources (e.g. coal mining areas, oil and gas fields) are presented.

Anthropogenic emissions of well-mixed greenhouse gases are currently the main drivers of tropospheric warming. Among the well-mixed greenhouse gases methane (CH4) and carbon dioxide (CO2) are the most important contributors. To limit the global warming, emissions of CH4 and CO2 must be reduced, and reduction claims need to be monitored. Additionally, knowledge of especially CH4 emission sources like landfills, oil, gas and coal production has to be expanded. Recently, a few different satellite and airborne instruments measuring anthropogenic greenhouse gas emissions at varying scales have been developed. The MAMAP2D-Light instrument developed and built at the Institute of Environmental Physics at the University of Bremen is a lightweight (~42 kg) single channel imaging spectrometer covering absorption bands of CO2 and CH4 between ~1575 and ~1700 nm with a spectral resolution of ~1.1 nm. The instrument is designed to fit into the under-wing pod of the Diamond HK36 TTC-ECO motor glider aircraft of the Jade University of Applied Sciences in Wilhelmshaven. At a typical flight altitude of ~1500 m the instrument samples 28 ground scenes across the ~600 m wide swath with a single ground sampling size of approximately 20 m across x 3 m along the flight track. While designed to detect and quantify CO2 and CH4 emissions from point sources, it additionally serves as precursor and demonstrator for the larger 2-channel imaging spectrometer MAMAP2D, which is currently being built. In this poster we present the results from a measurement flight targeting the CO2 emission plume of the coal-fired power plant Jänschwalde in Germany in June 2021, as well as initial calibration results.

As a recognized public health problem, air quality, especially surface concentration of air pollutants, is closely related to human living environment. Surface-level ozone is the focus and hotspot due to the trend of rising in recent years, and sufficient surface-level ozone data can provide reference for pollution control. This study aims to develop models for estimating high spatial resolution surface concentration of ozone based on TROPOspheric Monitoring Instrument (TROPOMI) data in China. By analyzing the source and sink of surface ozone, it is clear that ozone concentration in ambient air is influenced by background value, regional and local chemical generation, deposition, chemical removal and Interregional transport comprehensively, and then a formula for the calculation of surface ozone concentration is formed. Then, the Light Gradient Boosting Machine (LGBM) is used to integrate various satellite-based variables, numerical model-based meteorological variables and land variables in the above calculation formula to obtain the high spatial resolution surface concentration of ozone in China. In addition, three cross-validation methods (i.e., random, temporal and spatial) were used to verify the estimated surface concentrations. The results show that the correlation coefficient (r) can reach 0.7 and above, and the normalized root-mean-square-error (nRMSE) is basically 35% and below.

Under the umbrella of the international Commission on Atmospheric Chemistry and Global Pollution (iCACGP), the International Global Atmospheric Chemistry project (IGAC) endorsed the first phase of the Tropospheric Ozone Assessment Report (TOAR) in 2014. Its mission was to provide the research community and policy makers with a scientific assessment of the distribution and trends in ozone from the surface up to the tropopause. The first TOAR assessment, however, was not entirely conclusive and highlighted a list of scientific challenges to be addressed during a second phase. Among them is the difficulty to interpret tropospheric observations from space, especially when trying to interconnect data records from multiple satellites with significant differences in vertical sensitivity, resolution and spatial domain. Additional factors impeding the satellite interoperability are time-varying biases and the lack of harmonization between the different satellite datasets, e.g., regarding geophysical quantities, units and the definition of the tropopause. The overall result is an ensemble of (local) ozone distribution and trend estimates with a large dispersion, impeding firm assessments relevant for policy and science. In response to the harmonization needs revealed by TOAR, for the second phase of TOAR, the Committee on Earth Observation Satellites (CEOS) has therefore initiated a coordinated activity on tropospheric ozone assessments from space. In this framework, this work reports on the vertical harmonization of tropospheric ozone time series recorded by past and present satellites. First results are illustrated with tropospheric ozone data records from GOME-2, IASI, OMI and TROPOMI, based on two approaches: (1) a model is used to match prior profile (and prior constraint) information in given retrievals; (2) a method is applied that removes a priori information from atmospheric state profiles that are obtained through an optimal estimation retrieval. In both approaches, profiles are vertically integrated up to a common tropopause and spatiotemporally gridded afterwards. Remaining biases and uncertainty contributions are quantified using ozonesonde network measurements as a mutual reference.

Assessments of past and recent changes in stratospheric ozone concentrations use vertically resolved observations by satellite and ground-based instruments that span nearly four decades and cover the global scale. The small magnitude of the actual trend in the atmosphere implies that the stability of these observing systems should be better than about 2% per decade in order to detect trends. Ensuring such a level of stability poses a challenge to the teams that provide data records consisting of individual measurements as well as to the teams that combine these single sensor data into a multi-decade Climate Data Record. In view of the upcoming 2022 WMO/UNEP Ozone Assessment, we carried out a comprehensive analysis of the temporal stability of stratospheric ozone profile data records by seventeen limb and occultation satellite sensors and by the ground-based ozonesonde, stratospheric lidar and microwave radiometer monitoring networks (NDACC, WMO GAW, SHADOZ). It is an update and extension of earlier work that contributed to the SPARC/IO3C/GAW LOTUS and the 2018 WMO/UNEP assessments. Such an update is necessary given that several satellite sensors (e.g., Aura MLS, OSIRIS, ACE-FTS) operate many years beyond their design lifetime, often with reduced sampling capabilities and ageing detectors. Also, additional satellite sensors are included in our analysis (e.g., SABER, OMPS-LP, SAGE III/ISS) and revised versions of satellite (e.g., OSIRIS, MIPAS, SCIAMACHY, Aura MLS) and ground-based data records (e.g., ozonesonde). We present estimates of the stability of single profile data by limb/occultation sensors (Level-2) with respect to the ground-based networks. Our analysis was extended to evaluate the stability of aggregated single-sensor and multi-sensor data (Level-3), especially because the latter form the basis of most trend assessments. In many occasions our ground-based assessment of satellite data helps identify what sensor(s) are the main driver(s) of the observed trend differences between different merged data records. The availability of various complementary satellite sensors allows us to single out localised (time, space) inhomogeneities in the ground-based records as well. This, in turn, provides a better understanding of the difference in trends from different ground-based techniques. Our work clearly demonstrates that simultaneous coherent analyses of multiple complementary satellite and ground-based records are a prerequisite to constrain the stability of observing systems to the level required for trend assessments.

We present the MErged GRIdded Dataset of Ozone Profiles (MEGRIDOP) in the stratosphere with a resolved longitudinal structure, which is derived from data by six limb and occultation satellite instruments: GOMOS, SCIAMACHY and MIPAS on Envisat, OSIRIS on Odin, OMPS on Suomi-NPP, and MLS on Aura. The merged dataset was generated as a contribution to the European Space Agency Climate Change Initiative Ozone project (Ozone_cci). The period of this merged time series of ozone profiles is from late 2001 until the end of 2020. The monthly mean gridded ozone profile dataset is provided in the altitude range from 10 to 50 km in bins of 10° latitude x 20° longitude. The merging is performed using deseasonalized anomalies. The created MEGRIDOP dataset can be used for analyses which probe our understanding of stratospheric chemistry and dynamics. To illustrate some possible applications, we created a climatology of ozone profiles with resolved longitudinal structure. We found zonal asymmetry/structures in the climatological ozone profiles at middle and high latitudes associated with the polar vortex. At Northern high latitudes, the amplitude of the seasonal cycle also has a longitudinal dependence. The MEGRIDOP dataset has been also used to evaluate regional vertically-resolved ozone trends in the stratosphere, including polar regions. It is found that stratospheric ozone trends exhibit longitudinal structures at Northern Hemisphere middle and high latitudes, with enhanced trends over Scandinavia and the Atlantic region. This agrees well with previous analyses and might be due to changes in dynamical processes related to the Brewer-Dobson circulation.

The nadir viewing TROPOMI spectrometer aboard the S5p satellite since October 2017, has both high spatial resolution and daily coverage of the Earth. More than three years of GODFIT total ozone and OCRA/ROCINN CRB (cloud reflecting boundary) cloud fraction and height operational level 2 data (versions ≤ 2.2.x) are available that are combined to retrieve tropospheric ozone in two ways. We retrieve tropical tropospheric column ozone (TTCO) [DU] using the (CCD) convective cloud differential method CHORA (Cloud Height adjusted Ozone Reference Algorithm) and upper tropospheric ozone volume mixing ratios (TTO) [ppbv] using the (CSA) cloud slicing method CHOVA (Cloud Height induced Ozone Variation Analysis). These algorithms are based upon techniques developed by Ziemke et al. (1998, 2001). Temporal sampling of cloud/ozone data is not necessary anymore due to the high amount of daily S5p measurements. Data are spatially sampled on a 3° latitude/longitude grid with 2° step size to retrieve upper tropospheric ozone in the CHOVA algorithm. The retrieval results are used to calculate monthly mean volume mixing ratios in the Pacific sector to height adjust the CHORA above cloud column ozone (ACCO) to a fixed pressure level, here 270 hPa. The CHORA algorithm has also been optimized for the TROPOMI instrument. In a post-processing step, ACCO that is retrieved in the Pacific reference sector is interpolated and smoothed in time/latitude space to reduce data gaps and scatter in the daily ACCO vector fields. Daily total ozone is averaged in a small grid box with a latitude/longitude resolution of 0.5° to minimize errors from stratospheric ozone changes. All datasets have been successfully validated against the SHADOZ ozone sonde profiles with low biases (

As climate change accelerates, both the rate of occurrence and severity of bushfires, especially in Australia, continue to increase. Consequently, we posit that it is crucial to have automated mechanisms in place to quantitatively assess how they affect a key parameter in the consideration of climate change, which is the increased size of the ozone hole over Antarctica. We develop a model, based on a random forest ensemble algorithm that is a collection of decision trees, to understand exactly how these two aspects of the warming climate correlate. In particular, we take in parameters such as bushfire frequency, breadth, and severity. The output of the model is the number of square kilometers that the ozone hole would be predicted to grow based on the inputs provided. The primary purpose of this model is to provide a baseline for machine learning-driven investigations into the correlation between bushfires and the ozone hole.

The SATCROSS mission aims at measuring the water vapor content in the lower troposphere. In the basic configuration, a train of transmitting and receiving satellites is placed along the same orbit in such a way their line of sight passes tangentially to the Earth atmosphere, at altitudes below 10 km. A couple of very close radiofrequency (RF) K-band signals, emitted by a transmitting satellite to a receiving one, are attenuated while crossing the troposphere: such attenuations are then processed in order to recover two-dimensional water vapour fields on vertical sections of the troposphere itself. This mission concept requires proper payload and mission analysis. With regard to payload, solutions compatible with CubeSat platforms have been sought. This implies design is characterised by a strong miniaturization that is driven by the usage of commercial-of-the-shelf (COTS) components: in turn, it means reduced production costs and times. The purpose-built scientific equipment consists of an antenna and alternatively a transmitter or a receiver. All other satellite systems can be recovered in the CubeSat market. For the scientific antenna, gain is recommended greater than 30 dB: this requirement leads to a corrugated horn or a reflector antenna. The scientific transmitter needs a significant power amplification chain that is able to produce RF power level abundantly above 30 dBm. Since the two very close signals have to be acquired together and detected separately, the selectivity of the scientific receiver must be accurately elaborated. Both the transmitter and the receiver may be based on either analog (PLL, RSSI) or digital (FPGA) architectures. Preliminary verifications demonstrate these challenging performances are achievable. Mission analysis focuses, first of all, on the definition of orbital altitude. Link budget suggests a shorter distance between the satellites: therefore, a reduced altitude is preferred. Similarly, ionizing radiation effects analysis leads to the exclusion of altitudes too much above 400 km, so as not to overload the payload shielding. On the other hand, atmospheric drag may cause a too short mission duration if orbital altitude is below 400 km. In the end, the proper quote results in about 400 km. Simulations provided additional information on the orbit and constellation design. With reference to the former, inclination was chosen to cover specific areas at mid-latitude: sun-synchronous orbit results as suitable according to the trade-off between link duration and number of accesses. In order to reduce the revisiting time, some satellite networks have been investigated: up to four trains of satellites were distributed along one, two and four orbits which differ one another in true anomaly and right ascension of the ascending node. A further hint comes from orbital altitude, which matches the International Space Station (ISS) one: a viable possibility may consist in accommodating the scientific transmitter on board ISS, so that just the receiving satellites have to be properly arranged.

The measure of water vapor (WV) in the lower troposphere is a critical issue, which still leaves important margins for improvements of observational data quality, as necessary for instance to enhance the forecast performance of Numerical Weather Prediction systems.A novel technique developed in recent theoretical ESA studies, called NDSA (Normalized Differential Spectral Attenuation), is based on the measure of differential attenuation of two continuous tones transmitted at 18.8 and 19.2 GHz on a line of sight radio link.These studies proved the good capabilities of the technique in the estimation of the integrated water vapor (IWV), thus fostering further research efforts, including those aimed at developing and testing a demonstrator capable to acquire the first NDSA measurements. We recently designed and implemented a first demonstrator of such technique operating in a ground-to-ground link configuration. The first prototype was upgraded with the realization of an improved transmitter and receiver units to ensure more accurate measurements of IWV. We will report and discuss the results of this activity and the results of a new measurements campaign conducted with the last prototype.The poster presentation will also describe the activities envisaged in the near and mid-term for further advancement of the instrument prototype, such as, in particular, a ground-to-air measurement configuration. The aim of this is a feasibility study for the realization of a system based on the NDSA technique that satisfies the basic requirements laying the foundations for the ultimate ambition of a NDSA space mission deployment.

The atmospheric water cycle is strongly associated with global climate variations, especially in the tropical region (30°S - 30°N). However, the representation of the relation from models and observation results remains unknown. Here the water vapor observed during 2003 ~ 2014 over the tropical region is analyzed concerning large-scale circulation for 7 global climate models (CMIP6 framework) and the new global water vapor climate data records (CDR) generated within the ESA Water Vapor CCI+ project (WV_cci). In addition, the data from ERA5 are also employed as reference data. As the ESA WV_cci observes water vapor over the land area under clear-sky conditions and all-weather conditions without heavy precipitation in the ocean, the evaluation is conducted with land-sea separation. The observational diagnostic relies on the decomposition of the tropical atmosphere into large-scale dynamical regimes using the 500 hPa atmospheric vertical velocity w500 (in hPa/day) as a proxy. The ESA WV_cci and the CMIP6 data are then sorted according to dynamical regimes (intervals of 10 hPa/day) allowing to study the evolution of the regimes in terms of frequency of occurrence and is linked to water vapor variation. Although the inter-model spread lies within the interannual spread of the ESA CCI+_WV data, there are noticeable differences between the models and observations that are linked to large-scale dynamics

Normalized Differential Spectral Attenuation (NDSA) measurements from microwave (MW) satellite signals open new possibilities for retrieving water vapor (WV) fields and profiles along atmospheric transects, thus increasing their use in atmospheric science, from climate studies to operational meteorology. One of the key perspectives of this technology is the application to small satellites as Cubesats. WV is the only atmospheric gas that naturally varies to a great extent in space and time; it is an efficient greenhouse gas and it takes part in water transition phases, ruling great quantities of latent heat transfer. The WV content constrains cloud formation and precipitation, and more generally drives atmospheric dynamics. Its correct initialisation in numerical weather prediction models is consequently a critical aspect in meteorological forecasting, including data assimilation (DA) in limited area models. When dealing with high resolution numerical modelling, the DA aims at correctly activating smaller scale phenomena, thus enhancing the capability of predicting or nowcasting severe weather events. SATCROSS (Italian acronym for “Corotating satellites for the estimation of water vapor in the troposphere”) is an ongoing project funded by ASI (Italian Space Agency) and aimed at assessing the feasibility and impact of WV estimates retrieved by NDSA measures from corotating small LEO (Low Earth Orbit) satellites. In the framework of this project, we have started evaluating the capability of such new measurements to influence and possibly improve weather forecasts. Different meteorological scenarios have been considered in an Italian domain (including marine, plain and mountain areas), in order to evaluate the impact of different atmospheric phenomenologies on both WV estimate accuracies and on forecast performances, when assimilating the retrieved WV data. An Observing System Simulation Experiment (OSSE) architecture was set up to support WV retrieval experiments from NDSA synthetic observations and then to assimilate such measurements in model simulations, resembling operational runs. In the present work, we show the results of such numerical assimilation experiments, as available up to now. WV estimates are taken on transects, corresponding to annular regions in the common orbital plane of the corotating LEO satellites, as this region is swept by the MW links generated by the pairs of transmitters and receivers onboard such satellites. Also their assimilation account for the expected characteristics of the retrieval errors and of the sampling temporal and spatial distribution. This is still a work in progress, as different solutions could be applicable regarding the number of simultaneously sampled MW channels, the satellite number and orbits, the retrieval algorithms and assimilation procedures, so the work will show what experimented up to now and what we could infer from some analyses, looking at some possible different solutions. The reference code for the atmospheric model simulations is WRF (Weather Research and Forecasting Model) and the assimilation is achieved through variational methods, with a main focus on the 3D-VAR, as the method mainly used in most of the meteorological centres for regional forecasts, with resolutions capable to solve at least larger convective phenomena.

Recently, the Normalized Differential Spectral Absorption (NDSA) has been proposed to retrieve the tropospheric Water Vapor (WV) content. Since NDSA generates Integral WV (IWV) measurements, inversion methods are needed to achieve the spatial distribution of the WV. In this study, we investigate on the inversion problem related to the acquisition system studied in the SATCROSS project, that is a constellation of Low Earth Orbit (LEO) satellites, orbiting in the same plane and along the same direction, in which part of the satellites carries onboard transmitters (Tx) operating in the K and Ku bands and the remaining are equipped with receivers (Rx). By applying the NDSA approach along the Tx-Rx links, each of them crossing the troposphere, a set of IWV measurements can be achieved, whose inversion yields the WV content in the annular region contained in the orbital plane of the satellites. The specific geometry of the inversion problem at hand perfectly fits the so-called external reconstruction problem (ERP), proposed for the inversion of the Radon transform of a field when such transform is available only in the external part of a body. Even though the ERP was first introduced in the early ‘60s and further developed and refined in the following decades with applications to radiology and non-destructive testing, to the best of our knowledge it has never been applied to atmospheric remote sensing problems. The ERP solution is based on elementary functions in the Radon domain (IWV measurements) whose inverse transform, which lies in the space domain (the desired WV function), are known. Such elementary functions are proved to form a basis for the Radon and spatial domains. Finally, it can be shown the existence of spatial elementary functions that are mapped onto a null function in the Radon domain, that is a “null” (WV) space - not achievable from Radon (IWV) data - exists. In this study, we show the application of a slightly modified version of the ERP approach, referred to as ERTA (External Reconstruction Tomographic Algorithm), to the reconstruction of the WV field from NDSA measurements. In particular, we consider the peculiarities of our setup with respect to typical examples of ERP applications, that is: a very narrow annular area (i.e., a inner/outer radius ratio close to one) to be reconstructed; a very limited number of links on which the IWV is available, due to the complexity of deploying a highly populated constellation of satellites. A method to tackle the presence of the null space is also proposed, relying upon the assumption that the underlying function to be reconstructed is rather smooth. Simulations were carried out to demonstrate the effectiveness of the ERTA method for the reconstruction of the WV field in the context of the SATCROSS project. Encouraging preliminary results were obtained by using high resolution atmospheric data generated by a numerical weather prediction model as WV target field, both in the presence and absence of signal impairments.

Years ago, some of the authors proposed the application of the so called Normalized Differential Spectral Absorption (NDSA) technique - which allows one to retrieve the integrated water vapor (IWV) along a radio link established between a transmitter (Tx) and a receiver (Rx) - to a constellation of co-rotating (CO-ROT) satellites carrying various numbers of Tx and Rx and orbiting in the same Low Earth Orbit (LEO) plane. The objective was to retrieve the two-dimensional field of water vapor (WV) on a tropospheric annular region in the orbital plane perpendicular to the Earth surface and crossed by the ensemble of radio links established among the set of transmitters (on board the first set of satellites) and receivers (on board the second set of satellites) operating in the K and Ku frequency bands. The IWV estimates provided by the NDSA method are affected by the presence of Liquid Water (LW) along the radio link. Specifically, the impact of LW on the IWV estimates provided by NDSA in the 17-21 GHz frequency range is a positive bias which is proportional to the total Integrated Liquid Water (ILW) content along the radio link. The effects of the presence of ILW were already analyzed in previous works where the use of the spectral sensitivity measured in the 30-32 GHz range was proposed for estimating the path integrated LW, and possibly for correcting the IWV measurements. However, such an approach based on the spectral sensitivity comes out to be quite limited even when the signal-to-noise ratio of the received signals is relatively high. In this work, based on the research activities of the ongoing Italian Space Agency (ASI) funded SATCROSS project, we propose the use of the power ratio of the received signals at 31.8 GHz and at 16.9 GHz for directly estimating the path integrated LW. An end-to-end simulation tool for simulating the signals on the receiving satellites is used for analyzing the ILW estimation performance at global scale. The results are obtained using ECMWF atmospheric scenarios and WRF numerical weather prediction models. As a matter of fact, the ILW content appears to be linearly related to the power ratio and the simulations carried out under the hypothesis of realistic values of the signal impairments parameters show that the estimate of the power ratio is much more reliable than that of the spectral sensitivity and that the mass ILW amount can be estimated for values up to 4g/m2

Greece is located in the Eastern Mediterranean and exhibits favorable conditions for forest fires during the summer period. In the first two weeks of August 2021, Greece has suffered a series of wildfires that have burned a large part of the island of Evia and several areas of the Peloponnese (50,000 hectares). The present study aims at analyzing the impact of these fire events over Greece on atmospheric aerosol load using satellite data. Satellite information for aerosol height in the free troposphere can provide valuable information to the atmospheric modeling community by improving air quality forecasting and providing improved air quality and radiative forcing studies. Knowledge of the aerosol layer height is essential for understanding the impact of aerosols on the climate system and also can be useful for aviation and air quality alerts. Furthermore, the aerosol index can be used as an indicator for the presence of elevated absorbing aerosols especially for smoke and dust particles. TROPOMI onboard Sentinel-5P Aerosol Layer Height (ALH) and Aerosol Index (UVAI) products were analyzed for this work. GOME-2, onboard MetOp satellite platforms, can also be used to estimate the Aerosol height (AAH) and Aerosol Index (AI) over smoke scenes. Additionally, lidar profiles from the PANhellenic GEophysical observatory of Antikythera (PANGEA) can be used to verify the location of the aerosol layer in the atmospheric column and compare with the layer height retrieved by the satellite algorithms. Lidars is a valuable tool for aerosol characterization of biomass burning aerosols, since they can provide the vertical profiles of the aerosol optical and microphysical properties with high spatial and temporal resolution. A synergy of satellite data from GOME-2 and TROPOMI will be used to derive the aerosol optical and geometrical properties, during biomass burning events observed over Greece in the 2021 summer fires. The measurements allow us to monitor the evolution of fresh smoke aerosol, as well as to analyze its optical and geometrical properties (horizontal and vertical extend).

Abstract: The influence of aerosols on the energy balance of earth atmosphere radiation mainly depends on its aerosol characteristics, such as particle size distribution, chemical composition, and optical characteristics. The optical properties of aerosols have two basic parameters, aerosol optical depth (AOD) and single scattering albedo (SSA). SSA is the ratio of aerosol scattering coefficient to extinction coefficient, indicating the absorption of aerosol, and largely determines the positive and negative aerosol radiative forcing. However, there is still a lack of research on the hourly variation of aerosol absorption. Based on characteristics of high-frequency observations of geostationary satellites Himawari8, a new SSA inversion algorithm using the radiative transfer model parameterized by two-stream approximation is proposed in this paper. The prior parameters of the Ross-Thick Li-Sparse BRDF model come from MCD43C2 data, the values of hourly AOD come from the datasets generated by the optimal estimation method based on Himawari8 AHI data, and the prior knowledge of SSA and asymmetry factor comes from MISR data. We preliminarily obtained the SSA inversion results of China in April 2020 and verified with 7 AERONET sites data, showing that the N (total matchups) =156, R (correlation coefficient) = 0.38, and Q_0.05 (the percent of N that fall within the absolute difference of 0.05) = 55.6%. The obtained results will further help to understand the changes of aerosol absorption characteristics on an hourly scale.Keyword: single scattering albedo, Himawari8, Time-Space Sequence, Retrieval

Aerosol characterization is a key aspect in the design of algorithms for retrieving the CO2 dry-air mole fraction (XCO2) from satellite shortwave infrared (SWIR) observations. In the forthcoming Copernicus Anthropogenic Carbon Dioxide Monitoring (CO2M) mission, a high-resolution spectrometer for CO2 measurements (CO2I) will be operated simultaneously with a multi-angle polarimeter (MAP), which is expected to provide accurate information about aerosol size, single-scattering albedo and layer height. In principle, at least two approaches can be conceived for using the aerosol information from MAP measurements in XCO2 retrievals. One approach, currently being developed by SRON (Rusli et al., 2021), simultaneously fits MAP and CO2I measurements to retrieve aerosol properties alongside with XCO2. An alternative approach, being investigated by University of Leicester (UoL) and University of Lille/GRASP SAS, first retrieves aerosol properties from MAP measurements alone, and then uses the retrieved aerosol properties as input for an XCO2 retrieval only based on CO2I observations. In this presentation, the sequential approach to using MAP aerosol information in XCO2 retrievals will be discussed. The proposed approach uses the Generalized Retrieval Aerosol and Surface Properties (GRASP) algorithm (Dubovik et al. 2014) to infer aerosol properties from MAP measurements, and feeds its retrieved information to the UoL full physics CO2 retrieval, which is based on optimal estimation (Cogan et al., 2012). In order to evaluate the proposed approach, radiative transfer simulations were carried out in order to simulate CO2I and MAP measurements in an extensive set of realistic atmospheric scenarios. The synthetic MAP and CO2I observations were used in the sequential retrieval, and the retrieved XCO2 values were compared with the simulation input values. In the presentation, the simulation setup will be discussed, and a number of different setups of the GRASP aerosol retrieval will be compared in terms of their impact on the accuracy of the UoL XCO2 retrieval. Particular attention will be given to the problem of extrapolating MAP aerosol properties to the CO2I SWIR channels. Cogan, A., Boesch, H., Parker, R. J., et al. (2012), Atmospheric carbon dioxide retrieved from the Greenhouse gases Observing SATellite (GOSAT): Comparison with ground-based TCCON observations and GEOS-Chem model calculations, J. Geophys. Res.-Atmos., 117, D21301, doi: 10.1029/2012JD018087 Dubovik, O., Lapyonok, T., Litvinov, P., et al. (2014), GRASP: a versatile algorithm for characterizing the atmosphere, SPIE, doi: 10.1117/2.1201408.005558 Rusli, S., Hasekamp, O., aan de Brugh, J., et al. (2021), Anthropogenic CO2 monitoring satellite mission: the need for multi-angle polarimetric observations, Atmos. Meas. Tech., 14, 1167–1190, doi: 10.5194/amt-14-1167-2021

Aerosols along with tropospheric ozone and other reactive gases, are the main components characterizing the quality of the air we breathe, with implications for human health, life expectancy and the health of terrestrial and marine ecosystems. The Tropospheric Monitoring Instrument (TROPOMI) on board the Copernicus Sentinel 5 Precursor (S5P), the first of the European Sentinel satellites dedicated to monitoring of atmospheric composition, features a new aerosol product related to retrieval of the height of tropospheric aerosols (mainly desert dust, smoke and volcanic plumes). At present, daily global observations of aerosol height are not available on an operational basis. However, passive sensors, such as TROPOMI, can cover the entire earth in a single day. Even though the TROPOMI/Aerosol Layer Height (ALH) was released to the public on September 30, 2019 it is still a relatively new product, with validation still ongoing. The ground-based lidar observations of the vertical aerosol distribution provide the opportunity for important checks on the ALH data quality. The scientific community is making remarkable efforts in developing automatic ALH retrieval algorithms applied to lidar observations. In this study, we present results of a first validation analysis of the above-mentioned satellite-based product using independent, ground-based cloud-free observations taken at 18 sites in the frame of the European Aerosol Research Lidar Network (EARLINET) during September 2017- September 2021 and NATALI (Neural network Aerosol Typing Algorithm based on LIdar data) software (https://doi.org/10.5194/acp-18-14511-2018). EARLINET data was processed in a centralized way using the Single Calculus Chain (SCC), with specific configurations and settings; Level 2 data is compliant to all the QC procedures currently working on the ACTRIS/EARLINET database; also, the ACTRIS COVID-19 pandemic data collection https://doi.org/10.21336/gen.682q-8163 was used. Geometrical features of lofted aerosol layers (top and bottom) are calculated by applying the gradient method on the largest available wavelength of backscatter coefficient profile. Intensive optical parameters calculated from the backscattering, extinction and depolarization information (one hour averaged) within the layers with thicknesses of more than 300 m are considered further and analysed using NATALI software to determine the type of aerosol predominant in the layer. For overpass comparisons, ALH satellite pixels are average within 0.5 degrees radius around ground-based lidar location.

The retrieval of trace gas, cloud and aerosol measurements from ultraviolet, visible and near-infrared(UVN) sensors requires precise information on the surface properties that are traditionally obtainedfrom Lambertian equivalent reflectivity (LER) climatologies. The main drawbacks of using such LERclimatologies for new satellite missions are (a) climatologies are typically based on previous missionswith a significant lower spatial resolution, (b) they usually do not fully take into account the satelliteviewing dependencies characterized by the bidirectional reflectance distribution function (BRDF)effects, and (c) climatologies may differ considerably from the actual surface conditions especiallyunder snow/ice situations.In this work we present a novel algorithm for the retrieval of geometry-dependent effectiveLambertian equivalent reflectivity (GE_LER) from UVN sensors based on the full-physics inverselearning machine (FP_ILM) retrieval. The radiances are simulated using a radiative transfer modelthat takes into account the satellite viewing geometry and the inverse problem is solved using machinelearning techniques to obtain the GE_LER from satellite measurements.The GE_LER retrieval is optimized for the trace gas retrievals using the DOAS algorithm and thelarge amount of data of the new atmospheric Sentinel satellite missions. The GE_LER can either beused directly for the computation of AMFs using the effective scene approximation or a global gaplessgeometry-dependent LER (G3_LER) daily map can be easily created from the GE_LER under clear-sky conditions for the computation of AMFs using the independent pixel approximation.The FP_ILM GE_LER algorithm is applied to measurements of TROPOMI/S5P and it is being usedfor the operational retrieval of total ozone and cloud properties. The TROPOMI GE_LER/G3_LERresults for the fitting windows corresponding to O3, NO2, SO2, HCHO, H2O, and clouds are presentedand compared with climatological OMI and GOME-2 LER data.

Successful weather forecasts start from accurate estimates of the current state of the Earth system. Such estimates are obtained by combining model information with Earth-system observations via data assimilation. Cloud‐affected satellite radiance observations have been at the forefront of recent advances in data assimilation at ECMWF, particularly in the microwave where they provide information at a resolution of tens of kilometres, mainly on the distribution of rain and liquid cloud. Cloudy radiances in the visible contain a wealth of information on clouds, but have never been assimilated in global numerical weather prediction models. They are available at much higher horizontal resolution than microwave imagery, and are sensitive to the full depth of clouds in the atmosphere. While all-sky radiances at microwave frequencies are routinely included in the operational analysis, assimilating radiances in the visible part of the spectrum continues to pose many challenges. The reason lies in the fact that solar radiation originates from a single point in the sky, rather than from diffuse emission as in the case of microwave and IR radiances. This makes visible radiances much more sensitive to the details of the scattering phase function, and significantly adds to the computational cost of “observation operators”, i.e. the radiative transfer model to convert model profiles of atmospheric properties into radiances for comparing with the observed values. We are currently implementing the MFASIS (Method for FAst Satellite Image Synthesis) observation operator at ECMWF for monitoring cloudy reflectances. MFASIS, developed at LMU Munich, is a computationally-efficient model that allows the accurate simulation of cloud-affected visible satellite images, and is suited for operational applications in ECMWF's Integrated Forecasting System. MFASIS is based on a reflectance look-up table, and the state of the atmosphere is described by only a small number of parameters: the total optical depths and effective radii of water and ice clouds. The reflectance data are derived from the Ocean and Land Colour Instrument (OLCI) 665 nm radiances. OLCI is aboard two satellites, Sentinel-3A and Sentinel-3B, which orbit 140° out of phase with each other. The combined OLCI-A and OLCI-B instruments give good global coverage, with a short revisit time. The assimilation of visible satellite data will be extremely important for cloud reanalysis applications, and will open the door to unprecedented exploitation of past, present and future visible cloudy radiance observations. In this poster we will present the latest results of our work.

P. Litvinov1, O. Dubovik2, C. Chen1, D. Fuertes1, C. Matar1,B. Torres2, A. Lopatin1, T. Lapyonok2, M. Herreras1,2, Y. Derimian2, M. Herrera2, F. Ducos2, V. Lanzinger3, L. Bindreiter3, A. Hangler3, M. Aspetsberger3, C. Retscher4, A. Dandocsi4, D. Gasbarra4 1 GRASP-SAS, Generalized Retrieval of Atmosphere and Surface Properties, France2 Laboratoire d'Optique Atmosphérique, CNRS, Université Lille, France3 Catalysts GmbH, Huemerstraße 23, A-4020 Linz, Austria4ESA, ESRIN, Largo Galileo Galilei 1, 00044 Frascati (RM), Italy During a few last decades AERONET direct measurement and inversion products are the main validation dataset for satellite aerosol retrieval. Together with aerosol, Earth surfaces are the major contributor to the climate. Despite of evident need of the universal and robust reference dataset for surface, it still does not exist. Available ground-based measurements of the surface reflectance in worldwide location may not be universal tool for satellite surface product validation due to the fact that Earth surfaces, as a rule, are very spatially heterogeneous. Satellite surface products may suffer from aerosol contamination or lack of multi-angle measurements for BRDF characterisation. Their usage for surface validation always raises a question about accuracy and uncertainties. Essential enhancement of the surface retrieval accuracy, as well as its unification, can be achieved by inverting simultaneously AERONET ground-based and satellite measurements. This approach has already been implemented in GRASP retrieval developments in the frame of ESA GROSAT project. In this talk we present the concept of such combined synergetic retrieval and discuss the result of surface reference dataset obtained from diverse space-borne instruments like PARASOL, S5P/TROPOMI, S2 and S3/OLCI.

In remote sensing, the quantities of interest (e.g. the composition of the atmosphere) are usually not directly observable but can only be inferred indirectly via the measured spectra. To solve these inverse problems, retrieval algorithms are applied that usually depend on complex physical models, so-called radiative transfer models (RTMs). These are very accurate, however also computationally very expensive and therefore often not feasible in combination with strict time requirements of operational processing. With the recent advances in machine learning, the methods of this field have become very interesting in order to accelerate and improve the classical remote sensing retrieval algorithms. However, their application is not straightforward but instead quite challenging as there are many aspects to consider and parameters to optimize in order to achieve satisfying results. In this presentation we show a general framework for replacing the RTM used in an inversion algorithm with a deep neural network (DNN) that offers sufficient accuracy while at the same time increases the processing performance by several orders of magnitude. The different steps, sampling and generation of the training data, the selection of the hyperparameters, the training and finally the integration of the DNN into an operational environment are explained in detail. It is then demonstrated how the framework was used for the operational cloud product of S5P and S4 and how future improvements of the algorithms, by means of adding new physical models (e.g. ice-clouds), can be much easier integrated. Finally, the huge performance benefits of using a DNN instead of the original RTM also allow for new developments in the inversion part of the retrieval algorithms. Several ideas regarding this, e.g. global optimization techniques, are also shown.

Space-based atmospheric composition measurements, like those from Sentinel-5p TROPOMI, are strongly affected by the presence of clouds. Dedicated cloud data products, typically retrieved from measurements made with the same sounder, are therefore essential for atmospheric trace gas retrievals. Cloud products are used for masking scenes, filtering data, and as input to the modelling of atmospheric radiative transfer and the conversion of slant column densities into vertical column densities. The three main TROPOMI cloud data retrievals are: (i) L2_CLOUD OCRA/ROCINN CAL (Optical Cloud Recognition Algorithm/Retrieval of Cloud Information using Neural Networks; Clouds-As-Layers), (ii) L2_CLOUD OCRA/ROCINN CRB (same; Clouds-as Reflecting Boundaries), and (iii) the S5P support product FRESCO-S (Fast Retrieval Scheme for Clouds from Oxygen absorption bands for Sentinel), providing the following cloud variables: radiometric cloud fraction and cloud (top) height (all three retrievals) and cloud albedo/cloud optical thickness (ROCINN retrievals). These cloud variables are used in the retrieval of several TROPOMI trace gas products (e.g., ozone columns and profile, total and tropospheric nitrogen dioxide, sulfur dioxide, formaldehyde). The quality assessment of the cloud products and of the trace gas products is carried out routinely within the framework of ESA’s Sentinel-5p Mission Performance Centre (MPC), with ad hoc support from Sentinel-5p Validation Team (S5PVT) projects. We present the quality assessment of the three TROPOMI cloud products and the evolution of data quality through successive version upgrades, using ground-based and satellite data. Cloud height data from the three TROPOMI cloud products are compared to radar/lidar based cloud profile information from the ground-based networks ACTRIS-CLOUDNET and ARM. Deviations versus these networks are generally larger for high clouds, but with regard to tropospheric trace gas products S5P NO2 and S5P HCHO the quality of low cloud data is more critical. For TROPOMI NO2, a recent update (December 2020) of the FRESCO-S product led to a strong increase of the height of low clouds and subsequently a significant change in retrieved tropospheric NO2 improved the agreement with ground-based data. For TROPOMI HCHO, the update of OCRA/ROCINN_CRB had only a small impact on the cloud height of low clouds and therefore no large impact on the retrieved HCHO column. However, the sensitivity of TROPOMI HCHO to the cloud correction is demonstrated by the comparison with HCHO column retrievals from Aura OMI (QA4ECV HCHO data set). The discrepancies between both data sets are mainly due to the different cloud products: OCRA/ROCINN_CRB for TROPOMI HCHO, and CLDO2 for OMI QA4ECV HCHO. Radiometric cloud fraction and cloud height of the three TROPOMI cloud products and of OMI CLDO2 are compared, both at specific sites and globally. Discrepancies at several sites are reduced thanks to the product upgrades, e.g., due to a change in snow-ice treatment or surface albedo. In global comparisons, discrepancies between cloud products occur especially above snow-ice regions and when applying a less strict quality filtering.

FRESCO has been developed to retrieve cloud parameters from satellite spectrometers. FRESCO data are mainly used to correct cloud effects on trace gas retrievals, and to filter clouds in trace gas and aerosol retrievals. The FRESCO algorithm has been updated and implemented in the GOME-2 level 1 data processor (PPF) at EUMETSAT. The latest development is the FRESCO for Sentinels (FRESCO-S). There are significant changes in the latest implementation of FRESCO for S5P/S5. The main reasons for the changes are the high spectral resolution (narrower instrument spectral response function) and the variation of actual wavelength grid of TROPOMI/S5P. FRESCO-S products include effective cloud fraction, cloud height (cloud pressure), scene albedo, scene pressure. These parameters can be retrieved from the O2 A and B bands, with only the O2 A-band currently used for TROPOMI. The retrieval algorithm has been changed from Levenberg-Marquardt to optimal estimation. The surface directionally dependent Lambertian-equivalent reflectivity (DLER) derived from TROPOMI will be implemented in FRESCO-S for TROPOMI (GOME-2 DLER for S5), instead of the GOME-2 LER (used in current TROPOMI processor). Another improvement is about the a priori values for the cloud parameters. In FRESCO-S we use fixed a priori values for the scene parameters and cloud parameters. In the new version, we plan to use retrieved scene parameters as a priori values for the cloud parameters. In the presentation we will show the latest development of FRESCO-S and some preliminary results.

Breakthroughs in radar technology (Schottky diode-based frequency multiplied sources and solid-state transmitter) have recently fostered the construction of two G-band (frequency between 110 and 300 GHz) ground-based systems: the NASA-JPL VIPR system and the UKSA/CEOI funded GRaCE (G-band Radar for Cloud Experiments). The first measurements from these instruments confirmed the potential of G-band radars for sizing sub-millimeter crystals in ice clouds and raindrops in light rain. The presentation will discuss the deployment of a G-band radar in a space mission in the framework of Earth weather and climate monitoring programs.

Sentinel 5 - Precursor (S5P) is the first of the atmospheric composition Copernicus satellites and it was launched in October 2017. It carries one instrument onboard, TROPospheric Monitoring Instrument (TROPOMI), which makes it suitable for air quality, climate and greenhouse gases studies. New products have been developed through recent Sentinel - 5 Precursor+ Innovation (S5P+I) activities including retrieval of Aerosol Optical Depth (AOD) and Water Vapour (WV). Retrieval of AOD from S5P measurements is developed by GRASP-SAS. There are two algorithms employed, namely the Generalized Retrieval of Aerosol and Surface Properties (GRASP) and the Ozone Monitoring Instrument (OMI) Heritage algorithm, respectively. Both of them are using the UV-VIS spectral radiances from TROPOMI. One other Innovation activity is led by University of Leicester with their aim to retrieve dry-air mixing ratio water vapour and water vapour isotopologues using the shortwave infrared (SWIR) measurements from TROPOMI. The algorithm is based on the University of Leicester Full Physics (UoL-FP) processor which was previously used for the retrieval of XCH4, XCO2, HDO/H2O, planetary boundary layer (PBL) XH2O and solar induced fluorescence (SIF) from a number of satellite instruments operating in the SWIR spectrum. The algorithm has been adapted to use TROPOMI’s Bands 7 and 8 that cover the SWIR spectrum. Aerosol particles act as condensation nuclei in the moist atmosphere with the potential to take on water and form cloud droplets. Aerosols particles in the atmosphere absorb and desorb water vapour with a net forcing effect on the Earth’s radiation budget. They can also contribute to cloud properties modifications producing effects on the availability of precipitable water (e.g. the water cycle) and on the cloud lifetime. This study presents the synergetic use of two new TROPOMI products, xH2O and AOD, together with existing operational aerosol and clouds related products, namely Aerosol Index (AI), Cloud Fraction (CF), and Aerosol Layer Height (ALH) to determine the presence of Cloud Condensation Nuclei (CCN) and assess the aerosol’s hygroscopic properties using only TROPOMI data. Our study focuses on Europe as a region of interest for a period of no more than one year of data, i.e. 2019. The main purpose of this study is to present the feasibility of detecting cloud condensation nuclei using satellite data and to better understand the aerosol - cloud interaction impact on climate. Similar approach could also be used for future satellite missions (e.g. Sentinel 5, Sentinel 4).

Besides their strong contribution to weather forecast improvement through data assimilation in clear-sky conditions, thermal infrared sounders on board polar orbiting platforms are now playing a key role in monitoring changes in atmospheric composition. However, it is known that clear-sky observations are only a small part of the entire set of measurements, the remaining part are not being used as they are contaminated by either aerosols and/or clouds. Moreover, ice or liquid cloud retrieval of column and profile properties from passive and active measurements respectively help us in reaching a better understanding of climate processes. The information provided by the latter has allowed a significant advance in our knowledge of the vertical distribution of condensed water. However, it suffers from spatial coverage compared to passive measurements. It is therefore fundamental to better characterize cloud properties from passive measurements by using, for example, high spectral resolution instruments such as IASI and the future IASI-NG. The present study aims to quantify the potential and limits of thermal hyperspectral infrared sounders such as IASI or IASI-NG to retrieve ice cloud properties by using a representative dataset from the global operational short range forecast of the european center for medium-range weather forecast. We have made an information content analysis using Shannon's formalism to determine the level of the information about ice cloud properties in the IASI and IASI-NG spectra as a function of cloud water content. Based on this analysis, we have developped and tested an algorithm which allows to retrieve, from an optimal estimation approach, the cloud integrated ice water content and the cloud layer altitude. Using the IASI measurements over 2008, we show that our algorithm is able to reproduce the large scale structures of the distribution of integrated ice water content by comparison with climatologies from active sensors. We also discuss the capacities of the retrieval algorithm together with the ice cloud microphysical model to simulate the measurements made by IASI. We have taken into account the Signal-to-Noise ratio of each specific instrument and the uncertainties due to the non-retrieved atmospheric and surface parameters. The forward model is the fast radiative transfer model RTTOV which has been developped for satellite data assimilation in Numerical Weather Prediction models. The ice cloud microphysical model is based on the ensemble model of Baran and Labonnote (2007), where the bulk ice optical properties have been parametrized as a function of the ice water content (expressed in g/m³) and in-cloud temperature.

Atmospheric formaldehyde (HCHO) is a secondary product in the destruction of non-methane volatile organic compounds (NMVOC), through both natural and anthropogenic processes. With a relatively short lifetime of a few hours, the HCHO molecules are usually found close to their source and measuring HCHO from space is therefore highly relevant in obtaining information on NMVOC emissions and their role in air quality and climate. HCHO retrievals from satellite have so far be limited to polar orbiting sensors with a fixed local overpass time. This prevents the verification of the diurnal variation of HCHO and other NMVOC’s. The Geostationary Environment Monitoring Spectrometer (GEMS), launched on-board the GEO-KOMPSAT-2B satellite in February 2020 is the first geostationary sensor dedicated to air quality and atmospheric composition measurements. GEMS (observing South-East Asia hourly) will be joined by TEMPO in 2022 (United States) and Sentinel-4 (S4, 2023), monitoring Europe and Northern Africa. In the framework of the ESA-funded project “Sentinel-4 Level-2 Processor Component Development” (S4-L2), the S4-L2 team has reached out to the GEMS colleagues to jointly undertake the exploration of GEMS L1 and L2 data in preparation of the S4 mission. Geostationary sensors make fundamentally different demands on the HCHO algorithm as compared to polar sensors. In this work, we present first DOAS slant column retrieval results for HCHO from GEMS data and compare the results to those from the polar orbiting sensor Sentinel 5-P (S-5P) and the GEMS HCHO L2 product. The focus will be on the lessons learned from GEMS and how the algorithm developments performed in analyzing GEMS data are beneficial for HCHO retrievals from the future Sentinel-4. Particular attention will be paid on wavelength calibration aspects and to the schemes for the correction of background offset and zonal striping patterns in the retrieval results. GEMS data used in the presentation material were provided by NIER for validation and improvement research purposes as part of support for AO activities.

Arctic Amplification, the rapid increase of temperature at higher latitudes, is expected to impact various aspects of the Arctic system. Every polar spring, under the presence of sunlight, an intriguing natural phenomenon takes place: the so-called “bromine explosion,” a set of autocatalytic chemical chain reactions, which release bromine molecules into the atmosphere. These molecules are photolyzed to bromine atoms, which react with ozone, forming bromine oxides (BrO). Through this reaction cycle, ozone is massively depleted, and therefore the oxidizing capacity of the atmosphere (as ozone is the primary source of the hydroxyl radical, OH) potentially changes. Also, bromine oxidizes gaseous elemental mercury, transforming it into reactive gaseous mercury, which can enter and harm the ecosystem. Bromine molecules are released mainly from first year sea ice, while meteorological parameters are known to influence this process. Due to the complexity of the chemical reactions involved, the modeling of BrO is a challenging task. In this study, a novel machine learning approach using an artificial neural network (ANN) is presented in order to simulate enhanced tropospheric BrO in the Arctic atmosphere. The ANN uses sea ice age and meteorological drivers as input parameters. The 22-year satellite remote sensing dataset of Arctic tropospheric BrO developed in [1] is used to train and evaluate the ANN. Using only one year of observations as a training dataset, the ANN can reproduce the spatial variability of tropospheric BrO for many days of the 22-year time series. It is inferred from sensitivity tests employing the ANN that 2m air temperature and mean sea level pressure are the two input parameters that have the highest impact on the magnitude of simulated tropospheric BrO. The increasing trend reported in [1] is not seen in the simulations, potentially meaning that the trend is not driven by the input parameters chosen alone. Still, the ANN is a fast and computationally not demanding approach that can be integrated into chemical transport models as a sub-parameterization tool to predict the appearance of enhanced tropospheric BrO plumes quickly. This work was supported by the DFG funded Transregio-project TR 172 “Arctic Amplification (AC)3“. References [1]   I. Bougoudis, A.-M. Blechschmidt, A. Richter, S. Seo, J. P. Burrows, N. Theys, A. Rinke, Atmos. Chem. Phys., 20., 11869-11892, (2020).

Atmospheric Hydrogen cyanide (HCN) is one of the most abundant cyanides present in the global atmosphere and, due to its ability to influence the nitrogen cycle and satellite measurements of nitrogen oxides (NOx), understanding its physical and chemical nature are very important. Key processes driving tropospheric HCN variability are biomass burning, as the main source, and ocean uptake, as the main tropospheric sink. In the upper troposphere and stratosphere (UTS) the main HCN loss mechanisms are oxidation by hydroxyl radicals (OH) and by reaction with O(1D). The resulting HCN lifetime varies from 2–5 months in the troposphere to several years in the stratosphere. Due to its low chemical activity and its long lifetime, HCN is a good tracer of biomass burning, especially for peatland fires. Typically, peats contain a high concentration of partially decayed organic matter and once burned they are able to emit an incredible amount of carbon dioxide, particulate matter (PM) and other trace gases, including HCN, which affect regional air quality. Indonesia contains large peatlands that have been drained and cleared of natural vegetation, making them susceptible to burning. During 2015, one of the largest fire seasons in recent decades was observed in Indonesia due to the El Niño influence. In this work we present observations of HCN concentration over the Indonesia region measured by the Infrared Atmospheric Sounding Interferometer (IASI) instruments on-board the MetOp satellites during the Indonesian peatland fires from September to December 2015. Retrievals of HCN columns from the IASI measured radiances were made using the University of Leicester IASI Retrieval Scheme (ULIRS) adapted to that purpose. HCN columns are calculated on a 30-layer equidistant altitude grid from 0.5 to 29.5 km, using an optimal estimation approach for 8 channels in the range 710 - 717 cm-1 wavenumbers to include the strong HCN Q-branch at 712.5 cm-1. The Indonesia 2015 peat fire season case study is also investigated using an adapted version of the TOMCAT three-dimensional (3-D) chemical transport model (CTM). The HCN concentration over the Indonesian region has been modelled at a 2.8° × 2.8° spatial resolution from the surface to ~60 km for 12 idealised HCN tracers which quantify the main loss mechanisms of HCN, including ocean uptake and atmospheric oxidation reactions with O(1D) and OH. For the ocean loss we investigate the performance of different previously published schemes, while for the atmospheric loss we investigate the impact of using recent laboratory values rather than the recommended rates. The modelled HCN distribution has been compared with HCN columns measured by IASI and with the UTS profiles measured by the Atmospheric Chemistry Experiment Fourier Transform Spectrometer (ACE-FTS) over the region.

This work presents a 25-year multi-satellite data record of formaldehyde (HCHO) observations. Within the Quality Assurance for Essential Climate Variables (QA4ECV) project, a 20-year level-2 data record (1997-2018) of HCHO columns was reprocessed using state of the art European retrieval algorithms applied to four low-earth-orbit UV-VIS spectrometers: GOME/ERS2, SCIAMACHY/ENVISAT, GOME-2/METOPA and OMI/AURA. Those products are openly distributed via the QA4ECV website (www.qa4ecv.eu). In addition, operational retrievals from the TROPOspheric Monitoring Instrument (TROPOMI) on board of the Copernicus Sentinel-5 Precursor (S5P) platform since October 2017, rely on similar algorithms, which facilitates their inclusion in a climate data record. Retrieval algorithms have been homogenized to ensure optimal consistency between the historical QA4ECV dataset and the new TROPOMI operational products. However, despite these efforts of homogenization, fundamentals differences remain between the datasets, such as the spatial resolution, the overpass time, the sampling period, and possible instrumental degradation effects. Auxiliary datasets such as the cloud product, the surface reflectivity or the a priori profiles are also more difficult to harmonize over such a long time period. We present the status of the HCHO long-term data record, with the aim to assess the needs towards the creation of an atmospheric essential climate variable for ozone and aerosols precursor.

Bromine monoxide (BrO), one of the most dominantly observed indicators of reactive bromine species in the ozone catalytic loss process, can be measured by space-borne remote sensing. In this study, we generated a tropospheric BrO column dataset covering a period of 14 years from 2007 to 2020 using GOME-2 measurements from Metop-A/B with the framework of the AC-SAF project. The operational algorithm is based on the retrieval algorithm described by Theys et al., 2011. As BrO absorbs ultraviolet radiation with specific absorption features, total BrO slant columns are retrieved from satellite measurements based on the differential optical absorption spectroscopy (DOAS) technique. Total BrO slant columns are separated into their stratospheric and tropospheric contributions based on the climatology of stratospheric BrO profiles with total O3 and stratospheric NO2 columns as input data. In this retrieval dataset, some updates were included such as a new surface albedo database and an improved snow/ice flag compared to the scientific prototype algorithm. This algorithm is planned to be applied to TROPOMI measurements for the tropospheric BrO column retrieval, which makes it possible to extend the time series and monitor the spatiotemporal variability with a finer spatial resolution and a high signal to noise. Here, we present a case study of bromine explosion events in the Arctic using GOME-2 and comparable TROPOMI measurements and some validation results with ground-based measurements.

ADM-AEOLUS provides global and vertical high-resolution wind measurements in the troposphere and lower stratosphere. The measurements are ideally suited for studying atmospheric dynamics. So-called streamer events are of particular interest. The work presented here is part of the ESA-LISA project. The circumpolar wind band (jet stream) represents a meridional barrier for air masses, but also energy fluxes. So-called "streamer events" are an example of how this barrier can be disrupted. In fact, they thus limit the impermeability of this meridional diffusion barrier. During such events air masses from lower latitudes are irreversibly mixed into the circulation at higher latitudes with various consequences for atmospheric chemistry, energy and momentum balance. It is also assumed that increased atmospheric gravity waves and possibly also even infrasound can be generated by streamer events. There is broad agreement that they are the consequence of enhanced planetary wave activity/breaking, which modulate the jet stream and thus effecting the exchange of air masses and energy between the equator and the pole. Planetary waves are well-known to be major drivers of large-scale midlatitude weather in the troposphere and meridional ozone transport in the stratosphere. When planetary waves break, air masses are cut-off the jetstream. We present a simple but robust method based on wind measurements and related quantities (vorticity, divergence) that reliably enables the detection of streamers in large data sets. The work presented here is based on the ECMWF reanalysis data (ERA-5) as a pre-study for the later analysis in the ESA-project LISA. A case study is shown which clearly reveals that the spatio-temporal structure of a streamer is closely linked to planetary wave activity.

The Copernicus missions Sentinel-5p (S5p) and Sentinel-4 (S4) focus on atmospheric composition, trace gas and greenhouse gas retrieval for air quality monitoring. Covering the UV/VIS/NIR spectral region, it is also possible to retrieve basic cloud information, which is an important input parameter to an accurate trace gas retrieval because of potential albedo and/or shielding effects. Furthermore, a detailed knowledge about the cloud properties themselves is an important contribution in measuring and monitoring the Earth’s radiation budget, and hence, the impact of clouds on climatological applications. Operational cloud products have been provided by DLR for heritage instruments like GOME, SCIAMACHY and GOME-2 and are or will be also provided for the polar orbiting S5P (operational since 2018) and for the geostationary S4 (to be launched early 2024). In addition to these Copernicus missions, the DLR cloud algorithms have also been applied to the NASA Deep Space Climate Observatory (DSCOVR) mission using the EPIC (Earth Polychromatic Imaging Camera) instrument.The DLR cloud retrieval algorithms are called OCRA (Optical Cloud Recognition Algorithm) and ROCINN (Retrieval of Cloud Information using Neural Networks). In a first step, OCRA applies a UV/VIS broad band color space approach to determine a radiometric cloud fraction. This color space approach makes use of the assumption that clouds usually have a higher reflectivity than the surrounding surface and that the cloud reflectivity is almost wavelength independent across the UV/VIS region. In a second step, ROCINN retrieves cloud-top height, cloud optical thickness and cloud albedo from NIR measurements in and around the oxygen A-band. The cloud parameters retrieved by ROCINN are provided for two different cloud models: one which treats clouds more realistically as layers of scattering water droplets (clouds-as-layers, CAL), and one which treats clouds as simple Lambertian reflectors (clouds-as-reflecting boundaries, CRB). The current status of the algorithms is presented along with recent developments and improvements as well as application examples.

Arctic Amplification describes the recent period in which temperatures are rising twice as fast as the global average and sea ice and the Greenland ice shelf are approaching a tipping point. As a result, the Arctic’s ability to reflect solar energy decreases and absorption by the surface increases. A simple assumption would be that if the sea ice extent has been reduced, then the spectral reflectance at the top of the atmosphere - Rtoa - would have also decreased across the Arctic. On the other hand, Arctic reflectivity also largely depends on the presence of clouds, shielding the underlying surface, and on changes of their optical and physical properties. Thus, the assessment of trends of spectral reflectivity and cloud properties are essential to understand those forcings and feedbacks considered drivers of Arctic Amplification as well as the interactions between the components of the Arctic cryosphere. We observationally tackle the stated problem investigating changes of Rtoa at selected wavelengths making use of spaceborne measurements of the Global Ozone Monitoring Experiment (GOME onboard ERS-2 and MetOp A/B/C) and of the Scanning Imaging Absorption Spectrometer for Atmospheric Chartography (SCIAMACHY onboard Envisat) for the period 1995-2018. We complement this record with cloud properties and fluxes at top of the atmosphere and at the surface, inferred from measurements of the post-meridiem orbits of the Advanced Very High Resolution Radiometer (AVHRR onboard POES). Although the pan-Arctic reflectivity has decreased, the analysis of regional trends shows distinct areas where the reflectivity trends diverge. While darkening areas can be attributed to seasonal sea ice decline, an increase of Arctic brightness over sea ice free regions can be largely attributed to changes in the optical properties of clouds. Clouds increase the pan-Arctic reflectance by decreasing their ice and increasing their supercooled liquid water contents at polar temperatures. In the last twenty years, mixed-phase clouds have cooled the surface at low Arctic latitudes both in summer and spring while contributing to its warming over the circumpolar belt in spring. We also present a preliminary quantification of an observational instantaneous cloud feedback within the assumption of a climate having reached the long-term equilibrium state. Comparisons with models’ output suggests that a better description of the thermodynamics phase of clouds will be instrumental in reducing the intra-model spread of cloud feedbacks.

Three-dimensional (3D) cloud structures may impact atmospheric trace gas products from ultraviolet-visible (UV-VIS) sounders. We used synthetic and observational data to identify and quantify possible cloud-related bias in NO2 tropospheric vertical column densities (TVCD). In standard NO2 retreivals cloud effects are corrected using a Lambertian cloud model based on radiometric cloud fraction estimates and photon path length corrections based on absorption by oxygen collision pair (O2-O2 ) at 477nm or Oxygen A-band around 760nm. However, the impact of clouds is much more complex, involving unresolved sub-pixel clouds, scattering of clouds in neighboring pixels and cloud shadow effects, such that 3D radiation scattering from unresolved boundary layer clouds may give significant biases in the trace gas retrievals. Two synthetic cloud setups are considered: The first includes simple 2D clouds with various geometrical and optical thicknesses. This can be used to systematically investigate the sensitivity of the retrieval error on solar zenith angle, surface albedo and cloud parameters. The second includes realistic three-dimensional clouds from an ICON large eddy simulation (LES) for a region covering Germany and parts of surrounding countries. The scene includes cloud types typical for central Europe such as shallow cumulus, convective cloud cells, cirrus, and stratocumulus. The synthetic clouds were input to the MYSTIC 3D radiative transfer model and visible spectra for low-earth orbiting and geostationary geometries simulated. These spectra were analysed with standard retrieval methods and cloud correction schemes that are employed in operational NO2 satellite products. For the observational data the NO2 products from the TROPOspheric Monitoring Instrument (TROPOMI) were used while VIIRS provided high spatial resolution cloud and radiance data. The NO2 profile, cloud shadow fraction, cloud top height, cloud optical depth, solar zenith and viewing angles, were identified as the metrics being the most important in identifying 3D cloud impacts on NO2 TVCD retrievals. For a solar zenith angle less than about 40° the synthetic data show that the NO2 TVCD bias is typically below 10%. For larger solar zenith angles both synthetic and observational data often show NO2 TVCD bias on the order of tens of %. Specifically, for clearly identified cloud shadow bands in the observational data, the NO2 TVCD appears low-biased when the cloud shadow fraction > 0.0 compared to when the cloud shadow fraction is zero. For solar zenith angles between 50-60°, about 16% of TROPOMI pixels with high quality value NO2 TVCD retrievals, were found to be impacted by cloud effects larger than 20%. Several approaches to correct the NO2 retrieval in the cloud shadow were explored. The air mass factor calculated using a fitted surface albedo or corrected by O2-O2 SCD can partly correct the cloud shadow effects, however these approaches are limited to the cloud-free conditions. A parameterization approach based on the relationships identified from the sensitivity study was developed, and this can be used to identify data with a significant bias for the standard NO2 retrieval.

Clouds play a vital role in driving the water and energy budget of Earth’s climate system. Satellite records are invaluable tools for studying cloud-climate interactions, particularly over the remote oceans and polar regions. Using three long-term global satellite-based cloud and radiation datasets and two global sea ice and sea surface temperature datasets, we present an analysis on the trends and variability of cloud (e.g. cloud fraction, cloud optical thickness, cloud phase) and radiation (e.g. bottom of atmosphere shortwave and longwave radiative fluxes) properties over the Southern Ocean and their link to Antarctic sea ice. We also explore the trends and spatial patterns of clouds and radiation in these regions and their relationships with climate indices. The results reveal that during the sea ice growth season, sea ice concentration anomalies exhibit a significant negative correlation with total cloud cover anomalies while the correlation with net radiation flux anomalies is less evident. This implies the role of open water in facilitating cloud formation through moisture supplies, while the radiative impact on regulating sea ice storage is limited. During the melting season, sea ice concentration anomalies show a significant negative correlation with net radiation flux anomalies, but the correlation with total cloud cover anomalies is much weaker. The cloud-top temperature is found to be strongly correlated with the Southern Annular Mode. Despite these relationships, there is little consistency in the averages and trends of cloud properties among the three datasets. The reliability of these datasets related to the Southern Ocean, as well as the role of large-scale atmospheric circulations in driving the observed trends and variability of cloud and sea ice will be discussed.

Shadows of clouds are observed by TROPOMI as a result of its high spatial resolution as compared to its predecessors. These shadows contaminate TROPOMI's air quality measurements, because they are generally not taken into account in the models that are used for the retrieval. If the shadows are to be removed from the data, or if shadows are to be studied, an automatic detection of the shadow pixels is needed. We present the Detection AlgoRithm for CLOud Shadows (DARCLOS) for TROPOMI, which is, as far as we know, the first cloud shadow detection algorithm made for a spectrometer. DARCLOS raises potential cloud shadow flags (PCSFs), and actual cloud shadow flags (ACSFs). The PCSFs indicate the TROPOMI ground pixels that are potentially affected by cloud shadows based on a geometric consideration with safety margins. The ACSFs are a refinement of the PCSFs using spectral reflectance information of the PCSF pixels, and indicate the TROPOMI ground pixels that are confidently affected by cloud shadows. We validate DARCLOS with true color images made by the VIIRS instrument on board Suomi/NPP orbiting in close constellation with TROPOMI. We conclude that the PCSF can successfully be used to exclude any cloud shadow contamination from the TROPOMI L2 data, while the ACSF can be used to select pixels for the scientific analysis of cloud shadow effects.

Stratospheric chlorine dioxide (OClO) is formed in the reaction of ClO and BrO. While OClO is not directly involved in stratospheric ozone depletion, it can serve as a semi-quantitative indicator of stratospheric chlorine activation. As OClO has a strongly structures absorption spectrum in the UV, it can be measured from space by several satellite sensors. The existing record of nadir OClO observations from the GOME, SCIAMACHY, OMI and GOME2 sensors serves as a record of polar chlorine activation and its variability over time. In this project, a DOAS based OClO product was developed for the TROPOMI instrument on Sentinel-5 precursor. The aim is to create a data set to continue the existing long-term record and to bridge the time until Sentinel-5 data become available. The TROPOMI OClO product developed in the S5P+I OClO project shows excellent signal to noise ratio, small offsets and low scatter compared to data from all previous sensors. Chlorine activation in both hemispheres is clearly detected and shows strong variability in the Northern hemisphere during the observation period (2018 – 2021). In the Southern hemisphere, much less variation is observed but still differences between individual years are apparent. Validation of the TROPOMI OClO product with ground-based zenith-sky observations at eight stations shows excellent agreement. Further improvement in the comparison could be achieved after introduction of an objective offset correction of the ground-based data. Some systematic offsets remain in the data product in times and regions without activation but these are small in comparison to OClO columns observed in chlorine activation situations. Stratospheric chlorine dioxide (OClO) is formed in the reaction of ClO and BrO. While OClO is not directly involved in stratospheric ozone depletion, it can serve as a semi-quantitative indicator of stratospheric chlorine activation. As OClO has a strongly structures absorption spectrum in the UV, it can be measured from space by several satellite sensors. The existing record of nadir OClO observations from the GOME, SCIAMACHY, OMI and GOME2 sensors serves as a record of polar chlorine activation and its variability over time. In this project, a DOAS based OClO product was developed for the TROPOMI instrument on Sentinel-5 precursor. The aim is to create a data set to continue the existing long-term record and to bridge the time until Sentinel-5 data become available. The TROPOMI OClO product developed in the S5P+I OClO project shows excellent signal to noise ratio, small offsets and low scatter compared to data from all previous sensors. Chlorine activation in both hemispheres is clearly detected and shows strong variability in the Northern hemisphere during the observation period (2018 – 2021). In the Southern hemisphere, much less variation is observed but still differences between individual years are apparent. Validation of the TROPOMI OClO product with ground-based zenith-sky observations at eight stations shows excellent agreement. Further improvement in the comparison could be achieved after introduction of an objective offset correction of the ground-based data.

Chlorine dioxide (OClO) is a by-product of the ozone depleting halogen chemistry in the stratosphere. Although being rapidly photolysed at low solar zenith angles (SZAs) it plays an important role as an indicator of the chlorine activation in polar regions during polar winter and spring at twilight conditions because of the nearly linear dependence of its formation on chlorine oxide (ClO). We compare slant column densities (SCDs) of chlorine dioxide (OClO) obtained from measurements performed by the TROPOspheric Monitoring Instrument (TROPOMI) with meteorological data and CALIPSO Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) polar stratospheric cloud (PSC) observations for both Antarctic and Arctic regions. The observed OClO SCDs are generally well (anti-) correlated with the meteorological conditions in the polar winter stratosphere: e.g. the chlorine activation signal appears as a sharp gradient in the time series of the OClO SCDs once the temperature drops to values well below the Nitric Acid Trihydrate (NAT) existence temperature T(NAT). Also a relation of enhanced OClO values at the lee sides of mountains can be observed at the beginning of the winters indicating a possible effect of mountain lee waves on chlorine activation. OClO SCDs also coincide well with CALIOP measurements for which PSCs are detected. Very high OClO levels are observed for the northern hemispheric winter 2019/2020 with an extraordinarly long lasting and stable polar vortex being even close to the values found for Southern Hemispheric winters. In the extraordinary southern hemispheric winter 2019 a minor sudden stratospheric warming at the beginning of September was observed. In this winter similar OClO values were measured as in the previous (usual) winter till that event but with a 1 - 2 week earlier OClO deactivation.

The first observational dataset of vertically resolved global stratospheric bromine nitrate (BrONO2) distributions as derived from Michelson Interferometer for Passive Atmospheric Sounding (MIPAS)/Envisat infrared limb-emission spectra between July 2002 and April 2012 are compared to atmospheric model results. The altitude profiles of BrONO2 volume mixing ratios have been retrieved from averaged spectra based on the recently available final MIPAS level 1b-dataset (v8). All main modes of spatial and temporal variability of stratospheric BrONO2 in the observations are well replicated by simulations with the chemical climate model EMAC (ECHAM/MESSy Atmospheric Chemistry): the large diurnal variability, the low values during polar winter as well as the maxima at mid- and high latitudes during summer. Three major differences between observations and model results are observed: (1) a model underestimation of enhanced BrONO2 in the polar winter stratosphere above about 30 km, (2) higher modelled values than observed globally in the lower stratosphere (up to 25 km) most obvious during night, and (3) lower modelled concentrations at low latitudes between 27 and 32 km during sunlit conditions. (1) is explained by the model missing enhanced NOx produced in the mesosphere and lower thermosphere and subsided at high latitudes in winter. (2) could only be reproduced by an enhanced loss of BrONO2 through hydrolysis at stratospheric aerosol particles in the lower stratosphere, requiring, however, surface area densities hardly compatible with observations. The enhanced daytime signal in the observations (3) could only hardly be matched in the model runs by drastically decreasing the loss-rate through reaction with O(3P). The observations have also been used to independently derive the total stratospheric bromine content relative to years of stratospheric entry between 1997 and 2007. With an average value of 21.2±1.4 pptv of Bry at mid-latitudes where the modelled adjustment from BrONO2 to Bry is lowest, the MIPAS data are compatible with estimations of Bry derived from observations of BrO.

Halogen radicals can drastically alter the atmospheric chemistry. In the polar regions, this is made evident by the ozone depletion in the stratosphere (ozone hole) but also by localized destruction of boundary layer ozone during polar springs. These recurrent episodes of catalytic ozone depletion are caused by enhanced concentrations of reactive bromine compounds. The proposed mechanism by which these compounds are released into the atmosphere is called bromine explosion - reactive bromine is formed autocatalytically from the condensed phase. The spatial resolution of S-5P/TROPOMI of up to 3.5 x 5.5 km² allows for an improved localization and a more precise specification of these events compared to previous satellite measurements. Together with the better than daily coverage over the polar regions, this allows for investigations of the spatiotemporal variability of enhanced BrO levels and their relation to different possible bromine sources and release mechanisms. Here, we present tropospheric BrO column densities retrieved from TROPOMI measurements using Differential Optical Absorption Spectroscopy (DOAS). The advantage of our retrieval is its independence from any external input. We utilized a modified k-means clustering and methods from statistical data analysis to separate tropospheric and stratospheric partial columns, thereby relying only on NO2 and O3 measured by the same instrument. In addition, satellite pixels are distinguished by their sensitivity to the lower troposphere, and BrO slant column densities (SCDs) were converted to vertical column values by using airmass factors derived from O4 SCDs and reflectance data. Events from the polar springs in 2019 and 2020 were analyzed and compared to model data from a WRF-Chem simulation.

Glyoxal (CHOCHO) in the atmosphere originates from the oxidation of natural and anthropogenic non-methane volatile organic compounds (NMVOCs) but also from direct emissions during fossil fuel and biofuel combustion as well as biomass burning. With a short lifetime, enhanced glyoxal concentrations are found near NMVOC sources. Glyoxal tropospheric column measurements from space are therefore very useful to provide information on NMVOC emissions at the global scale. The BIRA-IASB scientific glyoxal algorithm has been further developed as part of the GLYRETRO project, one of the components of the ESA S5p+Innovation programme, which aims at further exploiting the capabilities of the TROPOMI instrument. This DOAS (Differential Optical Absorption Spectroscopy) algorithm has been applied to not only the full time series of the TROPOMI observations, but also to the heritage nadir sensors OMI and GOME-2A and -2B. We show that the use of a common algorithm to those different sensors leads to consistent glyoxal column fields with median differences typically less than 20%. Furthermore, comparisons have been performed with MAX-DOAS glyoxal measurements at different sites in Asia and Europe. Overall, satellite and ground-based instruments capture similarly the glyoxal variability. In general, the retrieved columns also agree in amplitude, except in specific conditions such as for example low sun elevation. TROPOMI glyoxal column data may be requested on https://glyretro.aeronomie.be Lerot, C., Hendrick, F., Van Roozendael, M., Alvarado, L. M. A., Richter, A., De Smedt, I., Theys, N., Vlietinck, J., Yu, H., Van Gent, J., Stavrakou, T., Müller, J.-F., Valks, P., Loyola, D., Irie, H., Kumar, V., Wagner, T., Schreier, S. F., Sinha, V., Wang, T., Wang, P., and Retscher, C.: Glyoxal tropospheric column retrievals from TROPOMI, multi-satellite intercomparison and ground-based validation, Atmos. Meas. Tech. Discuss. [preprint], https://doi.org/10.5194/amt-2021-158, in review, 2021.

Determining the height of a volcanic SO2 cloud after a volcanic eruption is a challenging task in UV satellite retrievals. The height is nevertheless the most important parameter required to forecast the movement of the volcanic cloud and to determine the total SO2 column and ejected SO2 mass, especially for local authorities and aviation safety applications. Retrieval algorithms developed so far performed direct fitting and optimal estimation techniques to determine the height information, which is hidden in the spectral signature. They are computationally very expensive and time consuming and can therefore not be applied in near-real time operational retrievals, especially not in current and future satellite UV instruments with high resolution and related high data amount. We have therefore developed a combined principle component analysis and neural network retrieval algorithm called ‘Full-Physics Inverse Learning Machine’ (FP_ILM), which performs an extremely fast (3ms per TROPOMI pixel) yet accurate (

Sulfur dioxide layer height (SO2LH) is key information to understand processes driving volcanic eruptions, assess the impact of SO2 emissions on climate and mitigate volcanic hazard for civil aviation. However, the determination of SO2LH from ground is notoriously difficult (if not impossible) in particular for strong and rapidly evolving eruptions. Satellite nadir observations with wide spatial coverage offer a viable source of information on SO2LH for spatially extended plumes. In the ultraviolet, space-based retrievals of SO2LH rely on the fact that, for large SO2 columns, the optical light path (AMF) becomes dependent on the absorption by SO2 (and thus on its vertical distribution), and the information on SO2LH can be inferred from the analysis of the measured radiances coupled with radiative transfer simulations. However, current algorithms are mostly sensitive to SO2LH for SO2 vertical columns of at least 20 DU. In this work, we present a new SO2LH algorithm and apply it to measurements from the high spatial resolution TROPOspheric Monitoring Instrument (TROPOMI). It relies on an iterative approach making use of a SO2 optical depth look-up-table. The particularity of this scheme is that it is a Covariance-Based Retrieval Algorithm (COBRA; Theys et al., 2021). This allows the background contribution of the measured optical depth to be treated in an optimal way, improving the sensitivity of the SO2LH retrievals for SO2 columns as lowas 5 DU. We demonstrate the value of this new data through a number of examples and comparison with IASI SO2LH estimates and back trajectories analyses. Theys, N., Fioletov, V., Li, C., De Smedt, I., Lerot, C., McLinden, C., Krotkov, N., Griffin, D., Clarisse, L., Hedelt, P., Loyola, D., Wagner, T., Kumar, V., Innes, A., Ribas, R., Hendrick, F., Vlietinck, J., Brenot, H., and Van Roozendael, M.: A Sulfur Dioxide Covariance-Based Retrieval Algorithm (COBRA): application to TROPOMI reveals new emission sources, Atmos. Chem. Phys. Discuss. [preprint], https://doi.org/10.5194/acp-2021-294, in review, 2021.

Usually, horizontal homogenous atmospheric properties are assumed for the analysis of satellite observations of atmospheric trace gases. While for most cases, this simplification causes only small to moderate errors, for the observation of volcanic plumes this assumption can lead to very large errors. 3D effects can become especially important for satellite observations with high spatial resolution like TROPOMI on S5P. Three different 3D effects are investigated in this study: a) geometric light path effects: the light path from the sun to the surface and that from the surface to the satellite might not both cross the volcanic plume; b) effects of horizontal light paths: light scattered into the FOV might originate from regions outside the volcanic plume and thus lead to a decrease of the absorption signal; c) saturation effects (for SO2): for narrow plumes the SO2 absorption signal can be strongly suppressed because most of the backscattered light is absorbed by SO2 itself. We investigate all three effects with the 3D Monte-Carlo radiative transfer model TRACY-2. We consider typical volcanic plumes and make simulations for observations directly above the volcano as well as in the horizontally oriented plume far away from the volcano. In order to quantify the associated errors we compare the results of the 3D simulations with those from simple 1D simulations. Finally, we give recommendations on how to best address 3D effects for the analysis of satellite observations of volcanic plumes.

Monitoring and modeling of volcanic aerosols is important for understanding the influence of volcanic activity on climate. Here, we applied the Lagrangian transport model Massive-Parallel Trajectory Calculations (MPTRAC) to estimate the SO2 injections into the upper troposphere and lower stratosphere by the eruption of the Raikoke volcano (48.29N, 153.25E) in June 2019 and its subsequent long-range transport and dispersion. First, we used SO2 observations from the AIRS (Atmospheric Infrared Sounder) and TROPOMI (TROPOspheric Monitoring Instrument) satellite instruments together with a backward trajectory approach to estimate the altitude-resolved SO2 injection time series. Second, we applied a scaling factor to the initial estimate of the SO2 mass and added an exponential decay to simulate the time evolution of the total SO2 mass. By comparing the estimated SO2 mass and the observed mass from TROPOMI, we show that the volcano injected 2.1±0.2\,Tg SO2 and the e-folding lifetime of the SO2 was about 13 to 17 days. The reconstructed injection time series are consistent between the AIRS nighttime and the TROPOMI daytime measurements. Further, we compared forward transport simulations that were initialized by AIRS and TROPOMI satellite observations with a constant SO2 injection rate. The results show that the modeled SO2 change, driven by chemical reactions, captures the SO2 mass variations observed by TROPOMI. In addition, the forward simulations reproduce the SO2 distributions in the first ~10 days after the eruption. However, diffusion in the forward simulations is too strong to capture the internal structure of the SO2 clouds, which is further quantified in the simulation of the compact SO2 cloud during late July to early August. Our study demonstrates the potential of using combined nadir satellite observations and Lagrangian transport simulations to further improve SO2 time- and height-resolved injection estimates of volcanic eruptions.

Bromine monoxide (BrO) is a halogen radical capable of influencing atmospheric chemical processes, in particular the abundance of ozone, e. g. in the polar boundary layer and above salt lakes, in the stratosphere as well as in volcanic plumes. Furthermore, the molar bromine to sulphur ratio in volcanic gas emissions is a proxy for the magmatic composition of a volcano and potentially an eruption forecast parameter. The high spatial resolution of the S5-P/TROPOMI instrument (up to 3.5x5.5km2) and its daily global coverage offer the potential to detect BrO and its corresponding ratio with sulphur dioxide (BrO/SO2) even during minor eruptions and for continuous passive degassing volcanoes. Here, we present a global overview of BrO/SO2 molar ratios in volcanic plumes derived from a systematic long-term investigation covering three years (Januar 2018 to December 2020) of TROPOMI data. We retrieved column densities of BrO and SO2 using Differential Optical Absorption Spectroscopy (DOAS) and calculated mean BrO/SO2 molar ratios for various volcanoes. As expected, the calculated BrO/SO2 molar ratios differ strongly between different volcanoes, but also between measurements at one volcano at different points in time, ranging from several 10-5 up to several 10-4. In our three-year study of S5P/TROPOMI data we successfully recorded elevated BrO column densities at 314 volcanic events and were able to derive significant (coefficient of determination, R² exceeding 0.5) BrO/SO2 ratios on 84 events at 14 different volcanoes.

Satellite measurement of volcanic gas emissions has been a key method in volcano monitoring for decades. Until recently however, it has been largely limited to explosive events and a few very high flux passive emitters. The 2017 launch of TROPOMI (the Tropospheric Monitoring Instrument), on board ESA’s Sentinel-5P satellite platform, was a step-change in the spatial resolution of satellite remote sensing and has brought about the possibility of measuring passive degassing emissions from both lower altitude and lower flux volcanic sources. The individual orbit operational products from TROPOMI have had a poorer S/N ratio than had been expected. A newer algorithm (COBRA) based on optimal estimation has significantly improved the noise level, resolving passive degassing from larger sources such as Mt. Etna, Italy. However, there are still numerous volcanoes that are known from ground measurements and direct observations to be passively degassing, that are still below the detection limit. Long-term averaging of satellite remote sensing measurements is an established technique for low-altitude, low concentration emission sources, both anthropogenic and volcanic in origin. There are now over 3 years of scientific-quality SO2 measurements from TROPOMI, allowing us to study the temporal evolution of even very small volcanic plumes. Soufrière Hills volcano, Montserrat, has been passively degassing since the end of the recent period of high activity in February 2010. Gas emissions have been monitored using the installed scanning DOAS network, giving a comparative SO2 flux, but the plume is not visible in daily imagery from either the operational or COBRA data sets. Given the volcano’s geographical location within the West Indies, the wind direction is to the West 85-90% of the year. This is advantageous as any gas will be blown in the same direction for most of the year, making the averaging technique more effective. Averaging is performed for Soufrière Hills for 3 monthly periods from June 2018 to Aug 2021. A clear plume is visible from almost all periods, with an average daily SO2 mass emission of 174 tonnes. This averaging technique, applied globally, would allow for an improved global volcanic SO2 flux inventory, as well as enhancing the monitoring of many volcanoes that were previously undetectable via satellites.

Aviation safety can be jeopardised by multiple hazards arising from natural phenomena, e.g., severe weather, aerosols/gases from natural hazard, space weather. Furthermore, there is the climate impact of aviation, that could be reduced. The use of satellite sensors, ground-based networks, and model forecasts is essential to detect and mitigate the risk of airborne hazards for aviation, as flying through them can have a strong impact on engines (abrasion and damages caused by aerosols) and on the health of passengers (e.g. due to associated hazardous trace gases). The goal of this work is to give an overview of the alert data products in development in the ALARM SESAR H2020 Exploratory Research project. The overall objective of ALARM (multi-hAzard monitoring and earLy wARning system; https://alarm-project.eu) is to develop a prototype global multi-hazard monitoring and Early Warning System (EWS), building upon SACS (Support to Aviation Control Service; https://sacs.aeronomie.be). After introducing the mechanism of ALARM EWS (origin and accessibility to data, visualisation, notification), this work presents alert data products creation. These products include observational data, alert flagging and tailored information (e.g., contamination of flight level – FL). We provide to the stakeholders information about the threat to aviation. Three different manners are produced, i.e., early warning (with geolocation, level of severity, quantification, …), nowcasting (up to 2 hours), and forecasting (from 2 to 48 hours) of hazard evolution at different FLs. Note that nowcasting and forecasting concerns SO2 contamination at FL around selected airports and the risk of environmental hotspots. This study shows the detection of 4 types of risks and weather-related phenomena, for which our EWS generates homogenised NetCDF Alert Products (NCAP) data. The first type is the near real-time detection of recent volcanic plumes, smoke from wildfires, and desert dust clouds, and the interest of combining geostationary and polar orbiting satellite observations. For the second type, ALARM EWS uses satellite and ground-based (GB) observations, and model forecasts to create NCAP related to real-time space weather activity. Exploratory research is developed by ALARM partners to improve detection of a third type of risk, i.e., the initiation of small-scale deep convection (under 2 km) around airports. GNSS data (ground-based networks and radio-occultations), lightning and radar data, are used to implement NCAP data (designed with the objective of bringing relevant information for improving nowcasts around airports). The fourth type is related to the detection of environmental hotspots, which describe regions that are strongly sensitive to aviation emissions. ALARM partners investigate the climate impact of aviation emissions with respect to the actual atmospheric synoptical condition, by relying on algorithmic Climate Change Functions (a-CCFs). These a-CCFs describe the climate impact of individual non-CO2 forcing compounds (contrails, nitrogen oxide and water vapour) as function of time, geographical location and cruise altitude. Acknowledgements: ALARM has received funding from the SESAR Joint Undertaking (JU) under grant agreement No 891467. The JU receives support from the European Union’s Horizon 2020 research and innovation programme and the SESAR JU members other than the Union.