Satellite observations of the atmospheric composition focus increasingly on the troposphere, for air quality and climate applications. Because the spatial variability of trace gases is much larger in the troposphere and in particular in the boundary layer, spatial resolution becomes increasingly important. Tropomi (Tropospheric Monitoring Instrument) on the Sentinel 5 Precursor satellite is a mapping instrument that combines a spatial resolution of approximately 3.5 x 5.5 km² at nadir with daily global coverage. Future missions will go to even higher spatial resolutions, for example CO2M (2x2 km²), the Nitrosat EE11 candidate (500x500m²) and the proposed TANGO (Twin Anthropogenic Greenhouse gas Observers) (300x300m²), while the swaths are generally smaller. To deliver the best possible data from these new instruments, not only the hardware has to be developed, but also the retrieval algorithms need to be improved. An important contributor to the retrieval uncertainty is the auxiliary data, such a surface reflectances and a-priori trace gas profiles. Furthermore, retrievals algorithms can be improved by making use of information from different sensors, for example for cloud corrections. While previous algorithms focused primarily on monitoring concentrations, new algorithms should also consider emission quantification of industries and cities as an important application area.

In this contribution we will present the approaches for high spatial resolution tropospheric NO₂ retrievals, which are currently being developed for the CO2M and TANGO missions. The retrieval algorithms include the use of neural networks to improve accuracy and to improve the computational performance. Results from tests on synthetic data as well as on a special zoom dataset collected with Tropomi during the commissioning phase, will be presented.

EUMETSAT has been operating the Global Ozone Monitoring Experiment-2 (GOME-2) on the Metop series of satellites since 2006. Next to operating the platforms and instruments themselves, the generation of the operational level-1B products, as well as the continuous monitoring of the instrument health and keeping track of the inevitable in-orbit degradation of the instrument optical and electronic modules is under the responsibility of EUMETSAT headquarters. Building on this, a long list of level-2 near-real time and offline products as well as derived data records is provided by the EUMETSAT Satellite Application Facility on Atmospheric Composition Monitoring (AC SAF).
While the lifetime of the Metop-A platform has come to an end in November 2021, after 15 years in orbit, Metop-B and -C are continuing to provide critical data all around the clock and the year.
In parallel, the preparation of the next generation of operational Atmospheric Chemistry missions is ongoing. The Copernicus UVN/Sentinel-4 instrument, to be launched on the Meteosat Third Generation - Sounder (MTG-S) in the first half of 2024, will be Europe's first Air Quality mission in the geosationary orbit and provide an unprecedented observation of the diurnal cycle of several critical trace gases, such as ozone, nitrogen dioxide, sulfur dioxide and many others. The polar orbiting Eumetsat Polar System – Second Generation A (EPS-SG A) platform, with a planned launch date around mid 2024, will carry the Copernicus UVNS/Sentinel-5 instrument, along with a group of high quality instruments for the sounding of atmospheric profiles and the observation of clouds and aerosols.
This presentation will focus on the state and health of the GOME-2 instruments in orbit and provide an update of the status of the ground segment and monitoring capapility development activities on the way for the future missions at EUMETSAT. Special attention is given to the in-orbit calibration strategies envisaged, aiming at maintaining high quality products throughout the lifetimes of the missions.

The Sentinel-4 (S4) mission focuses on monitoring of trace gas column densities and aerosols over Europe at high spatial resolution with an hourly revisit time, thereby covering the diurnal variation of atmospheric constituents.

Two complementary projects are working on the development of S4 Level 2 (L2) operational products:
• ESA S4 L2OP: O3 total and tropospheric column, NO2 total and tropospheric column, SO2, HCHO, CHOCHO columns, aerosol and cloud properties as well as surface reflectance.
• EUMETSAT AC-SAF: H2O and SO2 Layer Height.

In this article we present the status of the operational S4 L2 products being developed in the framework of ESA S4 L2OP and EUMETSAT AC-SAF.

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 involving improved cloud altitudes.

Operational retrievals of trace gas column amounts assume (near) cloud free conditions. However, the large pixel size of the satellite
instruments (for example the TROPOspheric Monitoring Instrument on Sentinel 5P, TROPOMI-S5P, is 5.5 km by 3.5 km at nadir) imply that
pixels may be contaminated by sub-pixel sized cloud(s). Furthermore, clouds in neighbour pixels may lead to in-scattering of radiation or
cloud shadow effects, both which are three-dimensional (3D) radiative transfer effects that may both decrease (cloud shadow) and increase
(in-scattering) the retrieved trace gas amount.

In a recent series of papers (Emde et al., 2021; Kylling et al., 2021; Yu et al., 2021) it has been shown for both synthetic and observational data, that for NO2 the impact of 3D clouds may result in a NO2 tropospheric column density bias on the order of tens of %. This is on the same order has the single pixel NO2 bias requirement (van Geffen et al., 2019). The NO2 retrieval bias depends on the NO2 profile. It is small for background conditions (low amount of NO2 in the lower troposphere) and is significant for polluted conditions (high amount
of NO2 in the lower troposphere) and pixels containing cloud shadows.

We present first results from the MIT3D project which 1) use a unique synthetic data set representing geostationary (Ultra-Violet and Near Infra-Red, UVN-S4) and low earth orbit (TROPOMI-S5P) instruments, to find and quantify the parameters that affect NO2 retrievals by utilizing artificial intelligence (AI); 2) use observational TROPOMI-S5P NO2 and Visible Infrared Imaging Radiometer Suite (VIIRS) cloud products to investigate the associations between these by maximal information-based non-parametric exploration (MINE) statistics; and 3) use the gained knowledge about which parameters and how they affect the NO2 retrieval, to devise a new cloud correction method to improve the standard 1D NO2 cloud correction.

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 and we present here the latest results and updates on the methodology. 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.