The Twin Anthropogenic Greenhouse Gas Observers (TANGO) mission is an ESA Scout initiative scheduled for launch in 2028, designed to monitor anthropogenic and natural greenhouse gas emissions from point sources at high spatial resolution. The mission consists of two 16U CubeSat platforms flying in tandem with a temporal separation of less than one minute, each carrying a pushbroom imaging spectrometer.  The first satellite is dedicated to the measurement of atmospheric CO₂ and CH₄ in the 1.6 μm spectral region, while the second satellite is optimized for the detection of NO₂ in the visible spectral range. Both instruments provide observations over a 30 × 30 km² swath with a ground spatial sampling of 300 × 300 m². TANGO is designed to survey more than 10,000 emission sources per year with a nominal revisit time of four days, focusing on emissions from sources with annual fluxes ≥ 2.5 Mt CO₂ and ≥ 5 kt CH₄.

In this study, we investigate the synergistic exploitation of collocated CO₂ and NO₂ observations to achieve an improved characterization of atmospheric emissions. The primary objective is to quantify the source emission ratio CO₂/NO₂ and to derive diagnostic parameters that elucidate an effective reaction rate associated with the transformation NO + O₃ → NO₂ + O₂. The tandem configuration of TANGO facilitates quasi-simultaneous measurements of both trace gases, thereby minimizing temporal variability in ambient atmospheric conditions between individual observations.

We evaluate the proposed methodology using dedicated microHH large-eddy simulations that incorporate realistic operational scenarios, including variable source strengths, temporal offsets between the two satellite overpasses, and heterogeneous spatial discretizations for CO₂ and NO₂. These simulations enable a quantitative assessment of the feasibility and accuracy of retrieving emission ratios and chemical parameters under conditions representative of actual measurement configurations. Subsequently, we validate the approach by applying it to ENMAP (Environmental Mapping and Analysis Program) satellite observations, thereby demonstrating its practical suitability for prospective TANGO measurements. The results underscore the potential of synergistically exploiting multispecies trace gas measurements to enhance emission quantification and to advance the characterization of atmospheric chemical processes.

The TANGO mission is a small satellite mission to be launched in 2028, under the ESA Scout Programme tapping into NewSpace to quickly deliver affordable and innovative science, as part of ESA’s FutureEO Programme, within a budget of 35M€ and a schedule of three years from mission kick-off to launch.

How to cite: Borsdorff, T., Krol, M., Veefkind, P., and Landgraf, J.: Synergistic Use of CO2 and NO2 for Emission Characterization: A Study Using TANGO Mission Simulations and ENMAP Data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11178, https://doi.org/10.5194/egusphere-egu26-11178, 2026.

Air pollution and climate change are two major global challenges that threaten sustainable development, and the tracking and mapping of atmospheric pollutants and greenhouse gases help to keep these two problems in check. Benefiting from its wide spatial-temporal coverage, satellite remote sensing is indispensable in the earth observation systems to provide measurements of atmospheric chemical species for decades.

Since 2008, when the Chinese second-generation polar-orbiting meteorological satellite Fengyun-3A(FY-3A) was launched, China has developed the capability to acquire the global atmospheric chemical components data from the space on daily basis. The three instruments aboard the FY-3A/3B/3C satellites, the Medium Resolution Spectral Imager (MERSI), Total Ozone Unit (TOU) and Solar Backscatter Ultraviolet Sounder (SBUS), enable the retrieval of aerosol optical depth (AOD), ozone total column and ozone vertical profile. Together with the subsequently launched FY-3D, FY-3F and FY-3H, they have established a global atmospheric composition dataset over fifteen years. The Hyperspectral Infrared Atmospheric Sounder (HIRAS) carried by FY-3D, FY-3E, FY-3F, FY-3H has been successfully used to retrieve the vertical profile of ozone and other trace gases during both day and night. Meanwhile, the Greenhouse-gases Absorption Spectrometer (GAS) onboard FY-3D and FY-3H realizes global carbon dioxide (CO₂) monitoring. The Nadir-viewing and Limb-viewing Ozone Monitoring Suite (OMS-N and OMS-L) aboard FY-3F detect stratospheric and tropospheric trace gases with higher quality. The Advanced Geostationary Radiation Imager (AGRI), embarked on the new generation geostationary satellites FY-4A, FY-4B and FY-4C, realizes the continuous measurement of aerosols since 2016. In addition, the Geostationary Interferometric Infrared Sounder (GIIRS), as the world’s first thermal infrared hyperspectral detector in geostationary orbit, is used to monitor trace gases with high temporal and spatial resolution, such as ozone, CO,NH₃, and HCOOH.

How to cite: Gao, L., Zhang, Q., Wang, Y., Wang, Q., Zhang, L., Bi, Y., and Zhang, X.: A Review on Chinese Fengyun Meteorological Satellites in Atmosphere Composition Monitoring, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9929, https://doi.org/10.5194/egusphere-egu26-9929, 2026.

In recent years, a constellation of hyperspectral infrared sounders has been successfully launched into LEO and GEO orbits on board China’s FengYun meteorological satellites. The Geostationary Interferometric Infrared Sounder (GIIRS) on board the FengYun-4 (FY-4) satellites scans the East Asian region every two hours. The Hyperspectral Infrared Atmospheric Sounder (HIRAS) on board the FengYun-3 (FY-3) series of satellites forms a constellation in dawn-dusk, mid-morning and afternoon sun-synchronous orbits. This provides six global thermal infrared observations per day, with equatorial overpass times of 5:30 am/pm (FY-3E), 10:00 am/pm (FY-3F) and 2:00 am/pm (FY-3H) respectively.

In the first half of this presentation, we will introduce the ozone products (total columns and profiles) from GIIRS which are retrieved from the 9.6 μm absorption band via optimal estimation algorithm. These retrievals have been rigorously validated against ground-based measurements, multi-satellite retrievals, and reanalysis datasets. Importantly, GIIRS exhibits peak vertical sensitivity in the upper troposphere/lower stratosphere (UTLS) region, providing unique capabilities for investigating stratosphere-troposphere transport (STT). In theory, FY-4B/GIIRS's 2-hourly ozone data can provide detailed information about ozone changes during STT events, enabling STT's impact on tropospheric ozone to be measured more accurately.

In the second half of this presentation, we will introduce the ammonia (NH₃) retrieval data products from FY-3E, the world's first operational meteorological satellite in a dawn-dusk orbit for civil use. FY-3E provides global observations twice daily, at around 05:30 and 17:30 local solar time. Using the optimal estimation method, we have retrieved daily global NH₃ maps from January 2023 to the present. Our retrievals reveal significantly elevated total columns of NH₃ during dawn and dusk in several major source regions. These regions exhibit spatial patterns and seasonal variability that are similar to those observed by IASI and CrIS. Notably, higher total columns are retrieved at dusk over some important source regions compared with mid-morning and afternoon observations, potentially due to more intense emissions and diurnal temperature variations. Combining these observations with data from mid-morning (e.g. IASI) and afternoon (e.g. CrIS) satellites will significantly enhance our understanding of the nitrogen cycle.

How to cite: Zeng, Z.-C.: Atmospheric composition observed from a constellation of LEO and GEO hyperspectral infrared sounders onboard FengYun satellites, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10065, https://doi.org/10.5194/egusphere-egu26-10065, 2026.

Atmospheric SO₂ plays an important role in air quality and climate. Satellite remote sensing enables continuous monitoring of SO₂ from volcanic and anthropogenic sources. The Ozone Monitoring Suite (OMS) onboard the Chinese FY-3F satellite, launched in August 2023, is a new hyperspectral UV–VIS instrument designed for atmospheric trace gas observations. Here we present the global retrieval of SO₂ columns from FY-3F/OMS nadir measurements using a Differential Optical Absorption Spectroscopy (DOAS) approach. Instrument-specific processing schemes, including solar spectrum selection, spectral soft calibration, and background offset correction, were developed to mitigate along-track striping and across-track asymmetry in the initial retrievals. The FY-3F/OMS SO₂ products are evaluated against TROPOMI SO₂ retrievals over clean oceanic regions, volcanic plumes, and anthropogenic emission areas. The results demonstrate good stability over clean regions (precision ~0.15 DU) and a clear capability to detect both volcanic and anthropogenic SO₂ enhancements. Remaining uncertainties are mainly related to detector non-uniformity and AMF. These results provide a first assessment of the FY-3F/OMS capability for global SO₂ monitoring.

How to cite: yan, H.: Retrieval of SO2 columns from FY-3F/OMS instrument observations, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2882, https://doi.org/10.5194/egusphere-egu26-2882, 2026.

Knowledge of the sulfur dioxide (SO₂) layer height (LH) is crucial to improve our understanding of volcanic events, and their atmospheric and climatic impacts. It is also essential to better constrain SO₂ emissions and ensure aviation safety. While SO₂ vertical column density (VCD) retrievals from UV nadir satellite observations are well established for decades, accurate determination of the SO₂ LH remains a major challenge. Existing spectral fitting algorithms are either time-consuming or lack precision and sensitivity, particularly for low SO₂ amounts in the upper troposphere–lower stratosphere.

Here, we present a new Level-2 product of SO₂ LH and VCD derived from the second UV spectral band (BD2) of the TROPOspheric Monitoring Instrument (TROPOMI) onboard the Sentinel-5 Precursor platform. The BD2 covers shorter UV wavelengths than the commonly used third UV band (BD3), providing stronger SO₂ absorption features that enhance the sensitivity of the retrievals, despite higher noise levels. These retrievals were performed using the Look-Up Table Covariance-Based Retrieval Algorithm (LUT-COBRA, [1]), which has been further developed and optimized for computational efficiency [2]. This algorithm was applied to the complete TROPOMI BD2 dataset spanning 2018–2025.

We analyzed both global and regional SO₂ variability, including specific volcanic events and degassing case studies. Compared to BD3, our approach demonstrates an improved sensitivity, precision, and accuracy, outperforming the current operational TROPOMI SO₂ product. In addition, validation with IASI thermal infrared measurements shows a relatively good agreement, confirming the reliability of the results. Our BD2 SO₂ product provides an unprecedented opportunity to monitor volcanic SO₂ emissions and their impacts over the past eight years.

[1] Theys et al., Atmospheric Measurement Techniques, 15(16):4801–4817, 2022.
[2] Fabris et al., Atmospheric Measurement Techniques, 2025. doi: 10.5194/egusphere-2025-4026.

How to cite: Fabris, L., Theys, N., Clarisse, L., Franco, B., Vlietinck, J., Yu, H., Brenot, H., Danckaert, T., and Van Roozendael, M.: Enhanced SO2 plume height retrievals from TROPOMI band 2 using a look-up-table COBRA approach over the full 2018–2025 timeframe, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9530, https://doi.org/10.5194/egusphere-egu26-9530, 2026.

The relocation of coal production has driven the expansion of the coal chemical industry and associated pollutant emissions in northwestern China, a region with sparse ground-based monitoring. Although data assimilation frameworks combining TROPOMI observations and chemical transport models are widely applied to infer NO_x and SO₂ emissions, their ability to resolve spatiotemporal variability is limited by smoothed priors and parameterized uncertainties, particularly where prior emissions are weak. Divergence-based approaches are computationally efficient but typically assume fixed lifetimes, failing to capture the pronounced variability of SO₂ lifetimes under changing atmospheric conditions. In this study, we employ a light weight method based on a model free mass conserving estimates (MCMFE) framework to quantify co-emitted NO_x and SO₂ emissions from four coal-based regions in northwest China for the period 2019 to 2020. The MCMFE-NO_x emission estimates including inclusion of explicit observational uncertainty, have been extensively evaluated and demonstrated to be robust in previous studies. Building upon this foundation, the present study improves the framework by introducing an iterative training strategy (IT-NO_x). IT-NO_x increases the number of valid grids by approximately 13.5%, corrects about 4.2% of grids with more physically reasonable estimates, and resolves severe underestimation in roughly 0.23% of grids. For SO₂, the approach is newly formulated around a five-term equation that integrates TROPOMI SO₂ observations with ERA5 wind fields, allowing the derivation of dynamic driving factors of SO₂ emissions, including lifetimes, transport distances, and diffusion rates. Rather than relying solely on “bottom-up” inventories to provide the SO₂ a priori, pseudo-priors for SO₂ used in this study are constructed by multiplying MEIC-derived SO₂/NO_x ratios with IT-NO_x emissions. Compared with directly using inventories as a priori, the daily pseudo-SO₂ framework based on IT-NO_x better captures realistic spatial variability of key driving factors and reduce the occurrence of extreme diffusion rates. The 20^th to 80^th percentile ranges of inferred lifetimes span from 5.2 hours to 14.8 hours, revealing seasonal region-specific energy-use patterns. Distinct weekday/weekend contrasts linked to two different emission sectors (transportation with residential activities, and coal plants) are also exhibited. Approximately half of coal-plant-dominated grids show modest lifetime differences, consistent with continuous operations, while transportation and residential dominated grids generally decline during weekends, due to increased private travel and tourism. Compared with the MEIC inventory, 84% of NO_x grids and 92% of SO₂ grids show higher emissions, with regional means of 0.82 ± 0.02 µg/m²/s and 0.52 ± 0.15 µg/m²/s, respectively. It is hoped that these findings will drive a new approach to SO₂ emissions estimation, one in which emissions are based consistently on remotely sensed measurements and associated uncertainties, especially in rapidly developing coal-based regions in northwest China.

How to cite: Lu, L., Qin, K., Cohen, J. B., Lolli, S., and Tiwari, P.: Satellite based inversion with NOx derived priors uncovers underestimated SO2 emissions over coal-based regions of China, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5664, https://doi.org/10.5194/egusphere-egu26-5664, 2026.

The regional CAMS (Copernicus Atmosphere Monitoring Service) analyses mainly profit from the assimilation of data from ground-based monitoring stations to obtain the best representation of the atmospheric state over Europe. Modern satellite missions such as Sentinel-5P and Sentinel-4, now provide high-resolution retrievals of atmospheric trace-gas columns. In order to assimilate such retrievals in EURAD-IM (European Air pollution Dispersion – Inverse Model), an observation operator is developed and tested. The challenge lies in the degrees of freedom how to distribute column values vertically across the different model levels. Furthermore, the assimilation of noise in the retrieval data poses risks to a meaningful representation of air pollutants in the model analysis. In this study, we assimilate nitrogen dioxide and sulphur dioxide TROPOMI retrievals in EURAD-IM, both with and without observations from the European ground-based monitoring network. The three-dimensional variational data assimilation technique is applied. Additionally, we test the potential of assessing emission with four-dimensional variational data assimilation. We compare and evaluate the simulation experiments using ground-based and airborne data. This demonstrates the added value of the satellite data assimilation. In this way, the EURAD-IM assimilation system is prepared for the upcoming hourly data of Sentinel-4.

How to cite: Lange, A. C., Franke, P., and Friese, E.: Benefits and challenges of assimilating Sentinel-5P TROPOMI retrievals into the regional CAMS analysis with EURAD-IM, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19407, https://doi.org/10.5194/egusphere-egu26-19407, 2026.

Hyperspectral observations from geostationary satellites provide detailed spectral information that is highly valuable for air quality monitoring, enabling improved characterization of aerosols and trace gases through their distinct spectral signatures. The Geostationary Environment Monitoring Spectrometer (GEMS) offers continuous hyperspectral measurements with high temporal resolution over the Asia–Pacific region, making it well suited for monitoring diurnal variations in atmospheric composition. However, the relatively coarse spatial resolution of hyperspectral geostationary sensors limits their ability to resolve fine-scale spatial heterogeneity in air pollution, especially in urban regions. This trade-off between spectral fidelity and spatial resolution remains a fundamental limitation of single-sensor satellite-based air quality monitoring. To address this challenge, this study develops a deep learning–based fusion framework that integrates hyperspectral radiance from GEMS with high-spatial-resolution multispectral observations from the Geostationary Ocean Color Imager-II (GOCI-II). A self-supervised learning strategy is used to improve the spatial resolution of GEMS Level-1C (L1C) radiance by using spatial patterns from GOCI-II. This makes hyperspectral super-resolution possible without needing high-resolution hyperspectral ground truth data. Validation against the original GEMS L1C data shows that the super-resolved radiance is very consistent in both space and time, with correlation coefficients (R) over 0.95 and normalized root mean square error (nRMSE) under 10%. The resulting super-resolved radiance preserves spectral information while providing substantially finer spatial detail than existing satellite products. Furthermore, the enhanced hyperspectral radiance is linked to surface-level air pollutant (e.g., PM₁₀, PM_2.5, and NO₂) concentrations through artificial intelligence-based models, demonstrating its applicability for high-resolution air quality monitoring. The proposed multi-satellite fusion framework highlights the value of integrating complementary satellite observations with data-driven approaches for urban-scale air quality analysis.

How to cite: Choi, H., Lee, S., Kim, Y., and Im, J.: Deep learning-based super-resolution of GEMS hyperspectral data using GOCI-II fusion: Advancing high-resolution air quality monitoring, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-20990, https://doi.org/10.5194/egusphere-egu26-20990, 2026.

For the first time, a Pandora spectrometer has been deployed in the central Himalayan region as part of the Pandora Global Network (PGN), at ARIES, Nainital (29.36°N, 79.36°E; 1970 m a.s.l.), a high-altitude remote site in South Asia where Pandora coverage has been negligible; however, pollutant concentrations across South Asia remain very high. Although the site is elevated, it is located adjacent to the Indo-Gangetic Plains (IGP) and is therefore affected by the transport of pollutants from the IGP.

The instrument retrieves column densities of key trace gases, ozone (O₃), nitrogen dioxide (NO₂), and formaldehyde (HCHO), including their total column (TO₃, TNO₂, THCHO) and lower-tropospheric column (LTrNO₂, LTrHCHO) amounts, suitable for validation of satellite observations in this complex mountain topology. Analysis of observations from January 2024 to June 2025 shows clear seasonality, with elevated springtime columns (TNO₂: 4-5 × 10¹⁵ molecules cm⁻²; LTrNO₂: 1 ± 0.1 × 10¹⁵ molecules cm⁻²; LTrHCHO: 0.8 ± 0.1 × 10¹⁶ molecules cm⁻²) and reduced values in the summer–monsoon period (LTrNO₂: ~0.3 ± 0.05 × 10¹⁵ molecules cm⁻²; LTrHCHO: ~0.2 × 10¹⁶ molecules cm⁻²). The seasonal cycle of column NO₂ agrees with surface in-situ NO_y, though their diurnal patterns differ: column NO₂ increasing steadily from morning until the evening period, while surface NO_y peaks around midday and column HCHO shows maximum value during daytime (12-13 hours IST), followed by a decline toward evening.

Pandora observations were also used to evaluate the performance of GEMS and TROPOMI satellite products for O₃ and NO₂ over a complex mountainous environment. For total column ozone, both GEMS and TROPOMI capture the day-to-day variability (R² = 0.98 for both satellites against Pandora). However, GEMS exhibits a systematic underestimation of about 15 ± 5 DU, while TROPOMI shows good agreement during the spring season but overestimates ozone by approximately 10 DU in other seasons. Similarly, both satellites represent the daily variability of TNO₂ reasonably well (R² = 0.71 for GEMS and 0.78 for TROPOMI). However, both instruments generally overestimate TNO₂. In contrast, the performance for LTrNO₂ is considerably poorer, with R² values of only 0.17 (GEMS) and 0.28 (TROPOMI). Thus, showing low sensitivity of satellite retrievals for NO₂ in the lower troposphere.  These results highlight the crucial role of high-quality ground-based measurements such as Pandora in validating satellite retrievals and advancing our understanding of trace gas behaviour in complex terrain.

How to cite: Kumar, M., Naja, M., Rawat, P., Srivastava, P., Tanimoto, H., and Crawford, J.: Assesssement of Trace Gases Variability and Satellite Retrieval Accuracy using Pandora Observations in the Central Himalayas, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-640, https://doi.org/10.5194/egusphere-egu26-640, 2026.

The first ever Geostationary Earth Orbit satellite air quality instrument over the Western Hemisphere,
Tropospheric Emissions: Monitoring of Pollution (TEMPO), has been scanning North America since August
2023 and providing the scientific community with hourly air quality observations. Algorithm science developed
for similar heritage instruments in Low Earth Orbit over the last two decades helped in the rapid development
and validation of key TEMPO products, such as nitrogen dioxide, formaldehyde, aerosol layer height, total
ozone, etc. Analyzing and demonstrating enhanced capabilities offered by TEMPO hourly observations rests
with the scientific community and the newly formed TEMPO science team. NOAA has been conducting scientific
work with TEMPO air quality products combined with NOAA operational GOES-19 Advanced Baseline Imager
air quality products to demonstrate their value for hazards monitoring and forecasting. This presentation will
showcase how NOAA is developing capabilities to verify emissions inventories in urban areas using TEMPO
nitrogen dioxide diurnal profiles, analyzing geography dependent pollutant exposure, and assessing how
pollution exposures at different times of day impact human health. NOAA’s goal is to use TEMPO data
operationally and help state and local agencies adapt to a new way of using satellite air quality data in their dayto-
day decision making processes.

How to cite: Kondragunta, S.: TEMPO Science and Applications, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-8371, https://doi.org/10.5194/egusphere-egu26-8371, 2026.

Human activities and climate change have profoundly changed emissions of air pollutants and greenhouse gases into the atmosphere. As countries move towards carbon neutrality and clean air, targeted emission control has become more important than ever to ensure rapid, deep and cost-effective emission mitigation. This ambition requires timely, high-resolution and accurate emission tracking, raising an unprecedented challenge to conventional emission inventories based on socioeconomic statistics and observation-based emission constraints that are subject to the resolution and coverage of observation data. In the advent of multi-satellite, multi-instrument, multi-species measurements of atmospheric constituents, together with rapid advancement of big Earth data and artificial intelligence techniques, a new paradigm of observation-based emission inversion becomes possible by strategically combining these sets of knowledge to guide a physics-based model framework in a computationally light manner. In this talk, starting from nitrogen oxides, we will present several scientific and methodological progresses to illustrate the emerging opportunity of this new paradigm for fast, fine, reliable satellite-based multi-species emission constraint, aiming to establish a comprehensive dataset to timely and accurately track air pollutants and greenhouse gases at fine scales.

How to cite: Lin, J., Kong, H., Wang, S., Tan, W., Wang, M., Xu, C., Zhang, Y., Hu, Y., and Shen, L. and the Atmospheric Chemistry & Modeling (ACM): Steps towards fast, fine, reliable satellite-based multi-species emission constraint, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2653, https://doi.org/10.5194/egusphere-egu26-2653, 2026.

Satellite observations of NO₂ play a central role in air quality and climate research; however, their quantitative interpretation is limited by uncertainties arising from retrieval algorithms, instrumental characteristics, and spatial representativeness. Robust interpretation of tropospheric NO₂ columns, therefore, depends on a comprehensive assessment of these uncertainty sources. Here, we investigate the primary contributors to uncertainty in TROPOMI NO₂ retrievals by examining individual retrieval steps and validating TROPOMI observations against independent Pandora and MAX-DOAS measurements. High-resolution chemical transport model simulations over Europe and the Netherlands are used to support and contextualize the analysis. Systematic biases are found in the stratosphere–troposphere separation of NO₂ in TROPOMI retrievals, with wintertime stratospheric columns overestimated by up to 0.15 Pmolec/cm² at high northern latitudes. These biases propagate into the tropospheric product, producing errors of up to 1.5 Pmolec/cm², primarily associated with limitations in the TM5-MP assimilation and further enhanced by large air-mass factor ratios under winter conditions. High-resolution LOTOS-EUROS simulations are used to evaluate representation errors associated with sub-pixel horizontal NO₂ gradients in satellite–ground-based comparisons, resulting in uncertainty estimates of approximately 6% at polluted sites. Differences in vertical sensitivity between TROPOMI and MAX-DOAS are shown to introduce substantial smoothing errors, reaching up to 20%. Comparisons between TROPOMI and Pandora direct-sun measurements reveal good seasonal agreement. Nonetheless, TROPOMI exhibits a negative bias relative to Pandora direct-sun measurements when using the default TM5-MP a-priori profiles. This bias is partially reduced by adopting higher-resolution CAMS-European a-priori profiles and further reduced when kilometre-scale simulations over the Netherlands are applied. These results highlight the critical importance of the spatial resolution of a-priori information in satellite–ground-based comparisons. Noticeable differences in both magnitude and seasonal variability are observed between MAX-DOAS, Pandora direct-sun, and Pandora sky-scan measurements, highlighting substantial intrinsic uncertainties within ground-based remote sensing products. Finally, uncertainty estimates derived from the distribution of differences between TROPOMI and ground-based observations generally exceed expectations based on the combination of individual uncertainty contributions, suggesting that current uncertainty estimates remain optimistic.

How to cite: Cifuentes, F., Eskes, H., Piters, A., Gomez, J., Douros, J., Pinardi, G., Friedrich, M., Dammers, E., Gebetsberger, M., and Boersma, F.: Characterizing uncertainty in TROPOMI NO2 retrievals across Europe with ground-based measurements and high-resolution modeling, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6019, https://doi.org/10.5194/egusphere-egu26-6019, 2026.

The Eastern Mediterranean and Middle East (EMME) is one of the most vulnerable regions to climate change globally and is becoming one of the world leading emitters of green-house gas (GHG) and air pollutants. Among these, nitrogen oxides NO_x (=NO+NO₂) are crucial to tropospheric chemistry, due to their role in the formation of tropospheric ozone O₃ and Particulate Matter (PM); both of which are harmful to human health and the ecosystem. NO_x are primarily emitted from the combustion of fossil fuels, which occurs in several sectors including transportation, energy production, industrial activities, residential heating, and agriculture. In spite of the direct and indirect threats of NO_x emissions, Saudi Arabia and the United Arab Emirates (UAE) continue expanding their fossil fuel production, with Saudi Arabia aiming to boost oil capacity to 13 million barrels per day by 2027, undermining its own 2060 net-zero pledge under the Saudi Green Initiative. The EMME region remains under studied regarding anthropogenic emissions, which highlights the need for accurate emission estimates to inform policy decisions.

In this work, we estimate NO_x emissions in the EMME region at a horizontal resolution of 0.5°, for the period 2019 to 2021. We employ the Community Inversion Framework (CIF) model, coupled to the CHIMERE chemistry transport model (CTM) and its adjoint, using a variational inversion method to construct NO_x emissions. We assimilate nitrogen dioxide (NO₂) observations from the TROPOspheric Monitoring Instrument (TROPOMI) on board the Copernicus Sentinel-5 Precursor (S-5P) satellite, and both anthropogenic and biogenic NO_x estimates from the Copernicus Atmosphere Monitoring Service (CAMS). Our emission data are close to those provided by other inventories. We examine key emitters in the EMME region, including countries that are affected by economic changes and/or political instabilities; such as Palestine, Israel, Lebanon, Iraq, Iran, Qatar, the UAE, and Saudi Arabia, among others. Our results show that, from 2019 to 2021, NO_x emissions exhibit a positive trend in most of the studied regions, except in Tehran (Iran) and Jeddah (Saudi Arabia), where we observe a decrease of NO_x emissions by -27% and -12% respectively. In the UAE, however, emissions increased by +17%, and in Yanbu (Saudi Arabia) by +24%, in 2021 compared to 2019. In Lebanon, a rise in NO_x emissions can be attributed to the country's economic crisis and shortages in national electricity supply, which led to a rapid increase in privately operated diesel-fueled energy producers. Our NO_x emissions data are expected to help policy makers monitor emissions in the EMME, at regional and national scales, to better tackle challenges specific to this region.

How to cite: Abeed, R., Fortems-Cheiney, A., Broquet, G., Pison, I., Berchet, A., Potier, E., Héraud, A., Rey-Pommier, A., Sciare, J., and Ciais, P.: Understanding NOx emission changes from 2019 to 2021 in the EMME region through variational inversions and satellite data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-5560, https://doi.org/10.5194/egusphere-egu26-5560, 2026.

Maritime transport is a pillar of the global economy, accounting for 75% of the European Union (EU) external trade, for example. It also has considerable environmental impacts. In terms of atmospheric pollution, shipping was responsible for about 39% of transport-related nitrogen oxide (NO_x) emissions in the EU in 2022. 

The ESA-funded Earth Observation for Ship Emission Monitoring project (EO4SEM) aims to provide shipping greenhouse gas and atmospheric pollutant emissions estimates that can support EU emission regulations. Specifically, it explores the potential of satellite-based Earth Observation to complement the bottom-up inventories that are driven by the automatic ship-tracking system called Automatic Identification System (AIS). Within this project, we have been developing atmospheric inverse modeling methods to derive estimates of NO_x emissions from sea shipping by processing Sentinel-5P/TROPOMI NO₂ images over European seas for the period 2019–2023. Different scales have been targeted and are discussed in this presentation. First, a sophisticated Bayesian atmospheric 3D chemistry and transport inverse modelling approach allows us to derive emission budgets for large sea areas. Second, lighter data-driven techniques derive emissions along individual shipping lanes on the one hand and instant estimates for individual large ships on the other hand. The AIS-driven bottom-up estimation model STEAM from the Finnish Meteorological Institute is used to support the analysis and then as a reference for the evaluation of the results.

Monthly NO_x emission maps at 0.5° resolution and corresponding budgets were derived over large sea regions defined by adapted International Hydrographic Organization (IHO) boundaries, using an inverse modeling approach based on the assimilation of TROPOMI NO₂ observations into the CIF-CHIMERE model.

The derivation of emission estimates along shipping lanes relies on the “divergence method”. This method is applied to individual TROPOMI images at the instrumental ground pixel scale. It derives corresponding NO_x emission maps. Accounting for the temporally-varying spatial coverage and noise of the quality-filtered retrievals from TROPOMI, we aggregate the results as monthly-mean NO_x lineic emissions (in kg/km/month). First comparisons between monthly-mean TROPOMI-based and STEAM lineic emissions estimates show a strong consistency for isolated, high-traffic lanes. However, the quantification remains challenging in complex areas characterized by high lane-intersection density.

Our estimates of instant emissions from individual large vessels are based on two types of approaches. Both involve the detection and inversion of the NO_x enhancement plumes downwind the moving vessels. The plume-detection algorithm is cross-referenced with the AIS information from STEAM to ensure that the high-concentration patterns identified in the TROPOMI images correspond to large ships, and to infer the trajectory of the latter. We use traditional point-source data-driven quantification methods: the cross-sectional flux and local divergence methods, that we adapted to account for the motion of the ship. The resulting estimates are confronted to the STEAM continuous estimates for individual ships, showing good consistency on average, but high uncertainty for the individual TROPOMI-based results.

How to cite: Diouf, M. M. N., Vignesoult, H., Fortems-Cheiney, A., Chevallier, F., Héraud, A., Beirle, S., Jalkanen, J.-P., Maragkidou, A., Brandão, F. G., Gini, R., Sengupta, D., Vitorino, J., Delavois, A., and Broquet, G.: TROPOMI-Derived NOx Emissions from Sea Shipping: Estimates Along Major Lanes and for Individual Vessels, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10642, https://doi.org/10.5194/egusphere-egu26-10642, 2026.

Satellite instruments such as TROPOMI are widely used for comprehensive global monitoring of nitrogen dioxide (NO₂). However, existing satellite retrievals only provide column densities (integrated trace gas concentrations) rather than surface concentrations, which limits their direct applicability for human-health studies

Over recent years, numerous machine learning models for the estimation of surface NO₂ have been developed. Such models use NO₂ vertical column densities (VCDs) from TROPOMI and ancillary input variables, such as meteorological data or bottom-up emission inventories, to predict surface NO₂ concentrations learned from in situ measurements. A consistent finding across studied is that land-use data and road networks are particularly helpful predictors, as they are available at street-scale resolutions and strongly linked to local NO₂ levels. However, their high spatial resolution introduces a major technical challenge: Representing square-kilometer-scale areas requires thousands of input data points, rendering many otherwise suitable neural network architectures, such as multilayer perceptrons, impractical to train. Consequently, previous approaches have relied on spatial aggregation methods, for example by computing coarse metrics such as road density at resolutions of 100 m × 100 m or coarser.

We develop a new methodology that processes such high-resolution ancillary data as images at street-scale resolution (~ 10 m × 10 m or finer), including road networks and building footprints from OpenStreetMap, detailed land-use information from the OSM Land-Use catalogue, and NO_x point sources from the European Release and Transfer Register (E-PRTR). A convolutional neural network is used to encode these high-resolution data into latent features. Combined with the TROPOMI NO₂ VCD and other low-resolution inputs, these are then used to estimate surface NO₂ concentrations via a multiplayer perceptron.

This approach is expected to

• improve predictive accuracy compared to models that rely on aggregation
• enable substantially higher horizontal output resolution down to the street scale
• provide a general framework for estimating surface concentrations of pollutants other than NO₂, as well as full diurnal concentration cycles

How to cite: Kuhn, L., Wagner, T., and Beirle, S.: Estimating street-scale NO2 surface concentrations from TROPOMI observations and high-resolution geographic data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4421, https://doi.org/10.5194/egusphere-egu26-4421, 2026.

How to cite: Jamal, S. A. and Batista, T.: Methodological Trends and Challenges in Deep Learning based Super-Resolution for Sentinel-5P Data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-7372, https://doi.org/10.5194/egusphere-egu26-7372, 2026.

Volatile organic compounds (VOCs) are key precursors of tropospheric ozone and secondary organic aerosols, a major component of PM_2.5, and several aromatic VOCs are toxic. Glyoxal is a short-lived oxidation product of many VOCs, yet global models consistently underestimate its abundance, indicating a substantial missing source. Here, we derive improved estimates of global biogenic, pyrogenic, and anthropogenic VOC emissions and new constraints on the atmospheric glyoxal budget, based on the first joint inversion of TROPOMI formaldehyde and glyoxal columns using the adjoint of the MAGRITTEv1.2 chemical transport model. The global NMVOC flux is estimated at 1070 Tg for 2021, 19% above bottom-up estimates, partitioned into 749 Tg from vegetation, 102 Tg from biomass burning, and 219 Tg from anthropogenic activity. Emissions of anthropogenic glyoxal precursors are 43% higher globally when constrained by satellite data compared with inventory-based simulations, with large underestimations in India, China, and Africa. The total glyoxal source is estimated at 100 Tg/yr, of which 41% originates from unidentified VOCs, predominantly biogenic and concentrated in the Tropics. Likely contributors include poorly represented formation pathway in isoprene oxidation under low-NO_x conditions and an underestimated contribution of monoterpenes. Validation against Pandonia Global Network, in situ, and MAX-DOAS datasets confirms improved agreement of the satellite-constrained model relative to the model based on inventory data alone.

How to cite: Stavrakou, T., Sfendla, Y., Müller, J.-F., Oomen, G.-M., Opacka, B., De Smedt, I., and Danckaert, T.: Global VOC emissions quantified from inversion of TROPOMI formaldehyde and glyoxal data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-2768, https://doi.org/10.5194/egusphere-egu26-2768, 2026.

The cloud-slicing retrieval technique has yielded new datasets of atmospheric composition in the free troposphere from satellite observations. The retrieval involves isolating clusters of satellite pixels above optically thick clouds from single overpasses or scans before regressing total column densities against corresponding cloud top heights to obtain a regression slope that is then converted to a single mixing ratio value representative of the average concentration of a target compound within the range of cloud top heights sampled. Recent datasets obtained with cloud-slicing include vertically-resolved concentrations of NO₂ and O₃ for multiple free tropospheric layers from TROPOMI and single-layer free tropospheric concentrations of NO₂ from TEMPO. Cloud-slicing for both NO₂ and O₃ suffers substantial data loss, as many clusters with non-uniform overlying stratosphere need to be discarded, due to the contribution of stratospheric variability to the regression slope. Cloud-slicing is yet to be tested on compounds sufficiently abundant in the free troposphere and without contamination from the stratosphere, namely formaldehyde (HCHO) and carbon monoxide (CO). Here, GEOS-Chem is used to generate pseudo-observations of HCHO and CO over target domains with distinct characteristics. Specifically, the remote troposphere (Pacific Ocean), and regions influenced by biomass burning (southern Africa) and anthropogenic pollution (South Asia). Cloud-slicing is applied to these pseudo-observations to tailor the retrieval steps to yield cloud-sliced mixing ratios that are consistent with the “true” mixing ratios as simulated by the model. According to preliminary data so far obtained for southern Africa in June-August, the peak of the burning season, lack of stratospheric contribution and greater data retention from cloud-slicing HCHO and CO total columns reduces noise in the cloud-sliced data, resulting in seasonal means that are more consistent with the "truth" than was possible with NO₂ and O₃. Cloud-sliced seasonal mean mixing ratios of HCHO and CO are typically within 5-10% of the “true” simulated mixing ratios and also achieve spatial consistency (R > 0.7). Though, cloud-sliced mixing ratios do underestimate large enhancements in HCHO and CO over the most intense biomass burning gridboxes. Work is underway to determine the cause for the bias over large sources, apply cloud-slicing to the other domains, explore the added value of free tropospheric HCHO and CO for understanding the oxidative capacity of the atmosphere, and quantify error contributions, including the representation error induced by sampling very cloudy scenes. Following this cloud-slicing characterisation, the algorithm developed with synthetic experiments will be applied to TROPOMI HCHO and CO data products to further extend the utility of Earth observations.

How to cite: Marais, E.: Using GEOS-Chem to design a cloud-slicing retrieval algorithm for application to TROPOMI formaldehyde and carbon monoxide, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-10717, https://doi.org/10.5194/egusphere-egu26-10717, 2026.

Emissions from wildfires have a large impact on the carbon cycle and air quality. Robust estimates of these emissions are important as they inform policy decisions. Bottom-up or top-down approaches are commonly used to estimate these emissions. Bottom-up inventories represent emissions as a product of a biome-specific emission factor and the amount of fuel burned. Top-down estimates utilize observations of trace gas from satellites or ground-based sensors to estimate emissions through an inverse modeling approach. A chemical transport model is used in conjunction with observations and a prior estimate (from a bottom-up inventory) to provide constraints on the emission sources. Bottom-up inventories typically have large uncertainties arising from variations in emission factors and discrepancies in the estimated mass of burned vegetation. While top-down estimates have the potential to mitigate these errors and provide more constrained emissions data, they still possess large biases and uncertainties. Atmospheric carbon monoxide (CO) is widely used as a tracer of wildfire emissions, and although various CO inversion studies have been conducted over the past two decades, there are still large discrepancies in reported top-down CO emission estimates. Here, we conduct a series of CO inversion analyses, focusing on the 2023 wildfires, to quantify the impact on the inferred CO emissions of the choice of data assimilation scheme employed, the specific observations being assimilated, the prior emissions inventory used, as well as the assumptions about the modeled chemical processes. Specifically, we compare the impact on the top-down emission estimates of using an ensemble Kalman filter and a four-dimensional variational data assimilation scheme to conduct the inversion. We also compare the impact of observations from the TROPOspheric Monitoring Instrument (TROPOMI) and the Measurement Of Pollution In The Troposphere (MOPITT) instrument, prior biomass burning emissions from the Global Fire Assimilation System (GFAS) and the Quick Fire Emissions Dataset (QFED), and examine the influence of the distribution of the modeled OH fields on the CO wildfire emission estimates. 

How to cite: Neeraj, H., Jones, D., Voshtani, S., Wunch, D., and Lutsch, E.: Assessing the sources of discrepancies in top-down CO emission estimates from wildfires, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-14807, https://doi.org/10.5194/egusphere-egu26-14807, 2026.

Carbon monoxide (CO) is an important air pollutant and precursor of ozone, plays an important role in atmospheric chemistry. Large quantities of carbon monoxide are emitted in coal-fired iron and steel  production processes, causing iron and steel plants to be the globally highest emitting point sources of CO. From the CORSO bottom-up global point source dataset, we select iron and steel plants thought to emit over 100 kt CO per year in 2021. For these plants, we obtain satellite-based emission estimates of carbon monoxide, applying the Cross-Sectional-Flux (CSF) method to TROPOMI for the period 2019-2021. We retain a global set of iron and steel plant yearly emission estimates of CO with satellite quantifications of 100 kt.yr^-1 and up. Previous research on European iron and steel plants  showed good agreement between bottom-up inventories, resource-intensive inversions and the low-cost CSF method. When we apply the CSF method outside of Europe we find substantial discrepancies between the bottom-up and top-down estimates, with large variations between regions. We examine potential sources of satellite under- or overestimation of the bottom-up inventory.

How to cite: Michaud van der Wal, C., Denier van der Gon, H., Leguijt, G., Guevara Vilardell, M., and Dellaert, S.: Quantification and evaluation of carbon monoxide emissions for high-emitting point sources using earth observation and bottom-up estimates , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-18003, https://doi.org/10.5194/egusphere-egu26-18003, 2026.

Ammonia (NH3), nitrogen oxides (NOx) and sulfur dioxide (SO2) are key precursors of secondary inorganic aerosols and strongly influence air quality, nitrogen deposition and ecosystem health. Yet bottom-up emission inventories and the temporal profiles for these species remain highly uncertain and often inconsistent across regions and sectors. Here we present a global dataset of anthropogenic NOx, SO2 and NH3 emissions over land, providing estimates of the daily 10-day mean emission fields since 2019 at 1.27° × 2.5° resolution. Emissions are derived from an atmospheric transport and chemistry inverse modelling system based on the global chemistry transport model LMDZ-INCA and a finite-difference mass-balance (FDMB) inversion approach. We account for satellite retrieval operators by consistently applying averaging kernels to the modeled NO2, SO2 and NH3 fields prior to model–observation comparison and emissions inversion. For NOx and SO2, we assimilate tropospheric NO2 columns from TROPOMI, with OMI-based NOx inversions used for consistency checks. For NH3, we use IASI total columns. The TROPOMI- and OMI-based NOx inversions show similar large-scale spatial patterns but differ regionally in magnitude, and generally indicate higher NOx emissions than bottom-up inventories over major source regions such as China and India. The TROPOMI-based SO2 inversions suggest lower anthropogenic SO2 emissions than bottom-up inventories at the global scale and across most major source regions, with global totals remaining relatively stable over 2019–2023. For NH3, the IASI-based inversion reveals persistent hotspots over South and East Asia—especially India and China—where inferred emissions exceed estimates from inventories, with pronounced seasonal peaks in high-emitting regions. Our dataset provides retrieval-consistent, time-resolved constraints on major aerosol precursors and implies systematic discrepancies between bottom-up inventories and satellite-constrained emissions over major source regions. The presentation details the methodological choices ensuring the routine estimates of such global maps of emissions, the relevance of their relatively high resolution, and investigations for a joint inversion of the three species to strengthen the consistency of the overall dataset.

How to cite: Li, F., Kumar, P., Broquet, G., Hauglustaine, D., Beaudor, M., Clarisse, L., Van Damme, M., Coheur, P., Cozic, A., Zheng, B., Li, H., Xu, J., Theys, N., Romero, B. R., Delavois, A., and Ciais, P.: Routine estimate of global 10-day mean maps of the anthropogenic NOx, SO2 and NH3 emissions over land since 2019 based on satellite observations , EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-9711, https://doi.org/10.5194/egusphere-egu26-9711, 2026.

Ammonia (NH₃) and nitrogen dioxide (NO₂) are key components of reactive nitrogen, with strong impacts on air quality, ecosystems, and nitrogen deposition. Long-term constraints on ammonia emissions and deposition remain uncertain due to sparse in situ measurements and limitations of individual satellite products. Here, we jointly assimilate five years (2018-2022) of NH₃ and NO₂ satellite observations over the Netherlands to improve constraints on reactive nitrogen concentrations, emissions, and deposition.

NH₃ retrievals from the Infrared Atmospheric Sounding Interferometer (IASI) and Cross-track Infrared Sounder (CrIS) are assimilated along with NO₂ observations from the TROPOspheric Monitoring Instrument (TROPOMI) within the Long Term Ozone Simulation-EURopean Operational Smog (LOTOS-EUROS) chemical transport model using a Local Ensemble Transform Kalman Filter (LETKF). The co-assimilation produces consistent year-to-year adjustments in modeled NH₃ concentrations, emissions, and deposition, reflecting the chemically linked nature of reduced and oxidized nitrogen. Our model results are evaluated against independent surface observations from the Dutch National Air Quality Monitoring Network (LML), showing reduced surface biases, improved correlations, and a clearer representation of diurnal variability. Sensitivity experiments demonstrate that including TROPOMI NO₂ alongside NH₃ observations leads to the lowest NH₃ surface biases, highlighting the added value of jointly assimilating chemically coupled species. Comparisons with the Dutch Measurements of Ammonia in Nature (MAN) network data show improved temporal correlations but persistent spatial biases related to representativeness differences, while MAN sensors co-located with LML stations exhibit consistent improvements.

In addition, synthetic NH₃ observations from the geostationary Meteosat Third Generation Infrared Sounder (MTG-IRS) are assimilated in a separate experiment to assess the potential of future high-temporal-resolution measurements. These experiments indicate that MTG-IRS will provide substantial added value for constraining ammonia emissions and deposition at diurnal scales. Our results demonstrate that co-assimilation of NH₃ and NO₂ satellite observations provides a robust pathway toward improved monitoring of reactive nitrogen and supports the design and exploitation of next-generation atmospheric composition missions.

How to cite: Wizenberg, T., Dammers, E., Segers, A., Herrera Gutierrez, B., Schaap, M., Shephard, M., Coheur, P., Van Damme, M., Eskes, H., Wichink Kruit, R., and van der Graaf, S.: Constraining ammonia emissions and deposition through joint NH3-NO2 satellite data assimilation in LOTOS-EUROS, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16526, https://doi.org/10.5194/egusphere-egu26-16526, 2026.

Quantifying ammonia (NH₃) volatilization in intensive agriculture remains challenging due to the high spatiotemporal variability of emissions. A key difficulty is to reconcile field-scale processes with the coarser resolution of satellite retrievals. To address this issue, this study proposes a robust framework to bridge the gap between top-down (TD) satellite constraints and bottom-up (BU) process estimates of NH₃ volatilization, and demonstrates its application to Queensland sugarcane belt over seven cropping seasons (2017–2023).

BU estimates were simulated using the NH₃ module of DNDC process-based model, driven by Sentinel-2 leaf area index (LAI) to infer fertiliser application timing, ERA5 meteorology data, and local agronomic nitrogen rate guidelines (the ‘Six Easy Steps’ program). TD emissions were derived from IASI v4 NH₃ concentrations using an upwind-ring background subtraction method and a steady-state mass-balance operator, with the lifetime diagnosed from regional GEOS-Chem simulations. Initial comparisons revealed significant discrepancies between these two estimates, with the original TD estimates exceeding BU estimates by 3.7 times (mean bias = 36 kg N ha^-1).

To reconcile these differences, we developed a bridging model that links TD and BU estimates as a function of meteorological conditions (temperature, ventilation) and fractional cane cover. These predictors act as a multiplicative correction and can effectively capture sub-grid source mixing and meteorological biases inherent in the satellite operators. Robust regression of the TD/BU ratio on these variables provides a statistically valid correction. Applying this adjustment reduced the normalized mean bias from 380% to 27%. The harmonized estimates are consistent with an independent estimate of regional NH₃ emissions of approximately 3.3 kt NH₃ yr^-1, confirming the dominance of diffuse agricultural sources in the region.

This framework yields more coherent NH₃ emission estimates for Queensland sugarcane and could in principle be adapted to other cropping systems where ground measurement data are sparse and satellite constraints are essential.

How to cite: Ma, Z., Pan, B., Parkes, B., Pang, A., Foster, T., and Lam, S. K.: Reconciling Top-Down and Bottom-Up Ammonia Emission Estimates over Queensland Sugarcane, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-310, https://doi.org/10.5194/egusphere-egu26-310, 2026.

Ammonia (NH₃) is an important atmospheric pollutant with environmental and public health impacts. In recent decades, NH₃ concentrations have increased due to intensified agricultural activities and industrial development, underscoring the need for high-resolution monitoring. However, sparse biweekly ground-based observations from the Ammonia Monitoring Network (AMoN) remain a major limitation for comprehensive spatiotemporal analysis. The United States (US) is a region where NH₃ monitoring is particularly important due to its extensive agricultural activities. In this study, we developed machine learning–based frameworks, including a deep neural network (DNN), random forest, and light gradient boosting machine, to estimate nationwide biweekly NH₃ concentrations and temporally downscale them to daily values across the contiguous US from 2017 to 2022. The models incorporate satellite-derived NH₃ column measurements, meteorological variables, land cover characteristics, livestock density, and AMoN ground-based observations. Among the tested approaches, the DNN demonstrated the strongest performance under both spatial cross-validation and independent testing, achieving a correlation coefficient of 0.79, a root mean square error of 0.98 µg m⁻³, and an index of agreement of 0.83. The model effectively reproduced fine-scale spatial variability in daily NH₃ concentrations at a 9 km resolution. Shapley additive explanations further revealed that temporally varying predictors—such as day of year and meteorological conditions—played a dominant role, alongside land cover and cattle density, supporting robust temporal downscaling from biweekly to daily scales. To assess spatial transferability, the framework was additionally applied to ground-based monitoring stations in the United Kingdom, where daily NH₃ observations are available, using leave-one-station-out and leave-one-year-out cross-validation schemes. Overall, our results demonstrate the potential of machine learning approaches to bridge temporal gaps in NH₃ observations and to generate high-resolution daily concentration estimates.

How to cite: Kang, E., Malik, S., Kang, Y., and Im, J.: Bridging temporal gaps: AI-based temporal downscaling of biweekly NH3 to daily scale with spatial transferability, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-19812, https://doi.org/10.5194/egusphere-egu26-19812, 2026.

Ammonium sulfate is a key component of secondary inorganic aerosols in northern China and contributes significantly to PM2.5 pollution. Through hygroscopic growth and enhanced light extinction, it also impacts atmospheric visibility and the regional radiative balance. In recent years, China’s clean air initiatives have significantly reduced surface PM2.5 concentrations. However, a decrease in total PM2.5 or bulk aerosol optical depth (AOD) does not necessarily imply a proportional, synchronised decline in ammonium sulfate. The lack of a long-term, interannually comparable record of ammonium sulfate aerosols hinders our ability to quantitatively understand the long-term changes in ammonium sulphate on a regional scale. Hyperspectral thermal infrared remote sensing offers a unique advantage in identifying the composition of aerosols. Ammonium sulfate exhibits resolvable absorption structures in the thermal infrared atmospheric window region, with a diagnostic spectral feature near 1115 cm⁻¹, which provides a physical basis for retrieving ammonium sulfate AOD.

In this study, we use long-term hyperspectral infrared measurements from the Infrared Atmospheric Sounding Interferometer (IASI) to construct an ammonium sulfate AOD time series for the North China Plain (NCP) from 2008 to 2025, and to characterise its spatial distribution, interannual variability and multi-year trends. Our focus is on the summer months, as ammonium sulfate over the NCP typically exhibits higher and more spatially continuous regional enhancement during this period. Additionally, infrared observations are sensitive to thermal conditions, and summer daytime provides more favourable conditions for achieving stable, interannually comparable results.

We use an optimal-estimation–based retrieval algorithm to retrieve ammonium sulfate AOD for clear sky observations. The state vector also includes interfering trace gases and surface temperature. The results show a significant decreasing trend in ammonium sulfate AOD over NCP during 2008–2025, with distinct spatial patterns and pronounced interannual variability. Furthermore, the retrieval results are compared with CAMS simulation products and long-term ground-based records of sulfate and  chemical composition. Overall, this work provides a satellite-based constraint on the long-term evolution of secondary inorganic aerosols over NCP. This offers new evidence with which to evaluate the effectiveness of mitigation measures and advance our mechanistic understanding of air pollution.

How to cite: Zheng, Y., Zeng, Z.-C., Clarisse, L., and Clerbaux, C.: Observed decadal variations of ammonium sulfate aerosols over northern China using the Infrared Atmospheric Sounding Interferometer (IASI), EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-6367, https://doi.org/10.5194/egusphere-egu26-6367, 2026.

Southern Africa is a climate-vulnerable region affected by biomass burning aerosols (BBA) emitted seasonally in central Africa, the primary source globally. By absorbing radiation and by deposition, these BBA have the potential of affecting in a very significant way both the regional radiative budget, the local meteorology and biogeochemistry, henceforth the regional climate. These impacts are governed by the high variability inherent to the fires and the short-lived atmospheric species.

In this study, we use satellite measurement of total column CO from three IASI instruments and the aerosol optical depth (AOD) by MODIS from 2007 to 2023, to investigate the climatology of transported BBA across the subcontinent at a daily timescale. We identify the seasonality, pathways and contribution of a meteorological phenomenon of the “rivers of smoke”, where the BBA plume is embedded in synoptic system, transporting high concentration of aerosol in the mid latitudes.

This intermittent transport happens seasonally, primarily above the continent, with a frequent second pathway along the western coast. The African regime, characterized by systematic fire and intermittent mid-latitude transport, contrasts with other important sources of tropical fire in the Amazon and southeast Asia. Our analysis reveals a maximum of contribution to the CO by the river of smoke in the southern Indian Ocean at 40°S. This pathway accounts for 25 to 30% of the regional CO concentration and peaks in September. In addition, the BBA, which are known to be in the fine mode, form plumes that contributes up to 60% of the fine mode fraction of AOD retrieved from MODIS in Namibia and South Africa, while the peak of contribution is in August, closer to the peak fire activity in July.

The river of smoke database from satellite observations established in this work provides a robust framework to tackle the variability of BBA.

How to cite: Guillemant, O., Cuesta, J., Gaetani, M., Pohl, B., Flamant, C., Dubovik, O., and Formenti, P.: Major systematic contribution of fine aerosols over southern Africa from rivers of smoke depicted from spaceborne multiyear observations, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11864, https://doi.org/10.5194/egusphere-egu26-11864, 2026.

Reliable surface monitoring of airborne particulate matter with an aerodynamic diameter smaller than 2.5 μm (PM₂.₅) remains limited in many regions around the world. In many countries, particularly developing regions, air-quality assessment relies on sparse or low-cost sensor networks with limited unit or data quality, or is even absent due to the high costs of installation and maintenance. Therefore, satellite observations are often proposed as an alternative option or complementary source for PM₂.₅ information. However, the extent to which satellite-based estimates can reliably represent surface PM₂.₅ concentrations relative to regulatory-grade ground-based measurements remains insufficiently quantified.

This study addresses this gap by evaluating satellite- and model-based PM₂.₅ estimates by comparing them with high-quality ground observations in the Kansai region of Japan. Kansai has a dense network of approximately 270 regulatory-grade PM2.5 monitoring stations, operated under Japan’s Air Pollution Control Act, in combination with urban, coastal, and topographic environments. That is why it is an ideal location and benchmark for conducting this study. Monthly PM₂.₅ observations for 2025 were used as reference data. This study utilizes model-based PM₂.₅ fields from the Copernicus Atmosphere Monitoring Service (CAMS) reanalysis. At the same time, satellite-derived aerosol information was retrieved from the Moderate Resolution Imaging Spectroradiometer (MODIS) Multi-Angle Implementation of Atmospheric Correction (MAIAC) aerosol optical depth (AOD) at 550 nm. The satellite- and model-based products were evaluated using the Pearson correlation coefficient, mean bias, and root mean square error (RMSE).

The results show that after quality control, 2,146 station–month pairs were available for evaluation. CAMS PM₂.₅ product shows only moderate agreement with ground-based measurements with Pearson r at 0.36. It shows a tendency to overestimate surface PM₂.₅ with a positive mean bias of 4.8 µg m⁻³ and an RMSE of 6.7 µg m⁻³. This result shows that CAMS captures large-scale variability, but does not fully represent local PM₂.₅ conditions at monitoring locations. By comparison, MODIS MAIAC AOD has a stronger correlation with observed PM₂.₅ (r = 0.56). This result indicates that changes in satellite-observed aerosol loading are more closely linked to variations in surface PM₂.₅. Using a simple linear model, AOD was able to explain about 31% of the monthly PM₂.₅ variability and substantially improved prediction accuracy, reducing the RMSE to 2.2 µg m⁻³. Seasonal analysis of the MODIS MAIAC AOD also reveals that there is a higher correlation with observed PM₂.₅ during winter and spring, with an r value of around 0.6–0.7. This is due to more stable atmospheric conditions and a lower boundary-layer height, allowing surface particles and atmospheric aerosols to be more closely linked. Meanwhile, in summer, the relationship weakened (r < 0.4), likely because a higher boundary layer and increased humidity reduce the direct connection between column aerosol measurements and surface PM₂.₅. Overall, these results provide quantitative evidence on the strengths and limitations of satellite and reanalysis products as alternative sources of air-quality information, particularly for regions with sparse or no surface monitoring.

How to cite: Handayani, A. and Higuchi, T.: Evaluation of Satellite and Reanalysis Products for Surface PM₂.₅ Using Regulatory-Grade Observations in the Kansai Region of Japan, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-16129, https://doi.org/10.5194/egusphere-egu26-16129, 2026.

We employ an updated retrieval of space-based methanol (CH₃OH) column measurements from the Infrared Atmospheric Sounding Interferometer (IASI) and an emission optimisation framework built on the adjoint of the MAGRITTE chemical transport model to assess terrestrial emissions of methanol to the atmosphere between 2008 and 2019. We first carry out a IASI CH₃OH validation study based on concentration measurements from three airborne campaigns over the U.S. in 2012-2013, using the model and the IASI averaging kernels to compute aircraft-based vertical columns directly comparable to IASI data.IASI is found to underestimate high columns and overestimate low columns in the considered region. A linear regression gives Ω_IASI = 0.46 Ω_airc + 10.6 · 10¹⁵ molec.cm^-2 , with Ω_IASI and Ω_airc the IASI and aircraft-derived vertical columns, respectively. Inverse modelling of terrestrial methanol emissions with the MAGRITTE model based on IASI columns corrected for biases using the above relationship leads to much-improved agreement over most regions against in situ observations from aircraft and surface measurement campaigns as well as column data at eight FTIR stations. The optimized global biogenic methanol emissions (160 Tg yr^-1 ) are 22-60% higher than previous top-down estimates, due to (1) column enhancements caused by the IASI bias-correction over source regions and (2) higher dry deposition velocities in the model over land, compared to previous model studies, based on a parametrisation constrained by field data from 13 campaign studies. The inversion results are less reliable over boreal forests due to shortcomings of both the bias-correction and the dry deposition scheme over these regions. The optimisation suggests large changes in the distribution and seasonality of biogenic emissions, such as enhanced emissions during warm and sunny periods over tropical ecosystems. In these regions, radiation and temperature seem to exert a stronger control on biogenic emissions than is currently accounted for in the MEGAN emission model, possibly due to leaf age effects currently not well accounted for in emission models.

How to cite: Muller, J.-F., Stavrakou, J., Franco, B., Clarisse, L., Amelynck, C., Schoon, N., Verreyken, B., Vigouroux, C., Mahieu, E., Makarova, M., and Strong, K.: Global atmospheric methanol emissions inferred from satellite IASI measurements and aircraft data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-4415, https://doi.org/10.5194/egusphere-egu26-4415, 2026.

Ozone and particulate nitrate, a key component of PM2.5, form through non-linear chemical interactions, with ozone formation governed by nitrogen oxides (NOx) and volatile organic compounds (VOCs), and particulate nitrate involving NOx (forming HNO3) and ammonia (NH3). This study utilise a multi-satellite observational approach to understand the formation response of ozone and particulate nitrate to their precursors over Delhi, world’s most polluted capital. We utilize Level-2 satellite data for NO2 and HCHO from TROPOMI (Sentinel-5P) and GEMS and NH3 from IASI (MetOp-B) over the spatial domain of 28°-29°N and 76.5°-77.8°E for 2023. Surface observations of ozone and PM2.5 were obtained from CPCB monitoring stations across Delhi after applying multi-level filtration. High resolution maps (1km × 1km) of NO2, HCHO, and NH3 along with their ratios (HCHO/NO2 and NH3/NO2) were generated, and their time series were extracted around each site locations (<10 km radius) to examine spatio-temporal patterns. Strong spatial and seasonal variability, with NO2 columns peaking at ~0.5 DU in winter and HCHO reaching up to ~0.85 DU in autumn is observed. HCHO/NO2 ratio shows VOC-limited and transitional ozone regimes during winter, which evolve into predominantly NOx-limited regimes during spring, summer-monsoon, and autumn. During spring, ozone concentrations peaks, ranging from 30–75 ppb, reflecting intense photochemical ozone production while time series of NO2 and HCHO columns around 0.2-0.3 DU, and 0.6-0.7 DU respectively. Ozone exceedance days (MDA > 50 ppbv) during spring shows elevated NO2 (~0.1 DU) and HCHO (~0.2 DU) levels, confirming photochemical production. Diurnal variation in NO2 and HCHO from GEMS, highlights seasonal and meteorological influence, with HCHO bias indicating a predominantly VOC limited regime.  The spatial gradient of NO2 (NO2/distance) highlights strong sinks in hotspot regions, particularly in winter and spring. A notable decline of 10-60% in ozone concentrations from spring to winter in these sink areas suggests substantial NOx-driven titration under low sunlight conditions. Site classification into urban, rural, and highway-proximal (within 500 m) categories shows consistently higher NO2 and HCHO levels in urban areas across all seasons, followed by highway sites. For particulate pollution, particulate nitrate, a secondary inorganic aerosol was found to significantly contribute to PM2.5 across seasons. PM2.5 levels peaked in autumn at all sites, followed by winter. Binned HCHO averages were higher in autumn, aligning with PM2.5 peaks and suggesting a biogenic contribution during extreme pollution events. Conversely, elevated NO2 during winter points towards a dominant inorganic and anthropogenic influence on PM2.5 enhancement in form of particulate nitrate.

How to cite: Tomar, V., Naja, M., Rawat, P., Sun, K., Kumar, R., and Kumar, U.: Chemical Signatures and Sensitivity of Ozone and Particulate Nitrate from Multi-Satellite and Ground Observations over Northern India, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-946, https://doi.org/10.5194/egusphere-egu26-946, 2026.

Tropospheric ozone is an important atmospheric trace gas that affects air quality, human health, and climate. However, its accurate retrieval from satellite observations remains challenging while in situ vertical profile measurements are sparse. The retrieval of tropospheric ozone is impeded by its weak signal compared to the dominant stratospheric ozone column, and current full physics retrievals often suffer from limited vertical resolution and show insufficient agreement to in situ observations. Recent advances in satellite instrumentation and machine learning provide an opportunity to overcome these limitations. In particular, the Tropospheric Monitoring Instrument (TROPOMI) offers high spatial resolution and signal-to-noise ratio, enabling more detailed daily observations of atmospheric ozone variability with global coverage.

We explore the feasibility of a deep-learning-based technique for retrieving high-resolution tropospheric ozone profiles from TROPOMI spectral measurements combined with auxiliary meteorological information, while avoiding some of the simplifying assumptions made in existing full physics retrieval approaches. Artificial neural networks are well-suited for this task, as they can learn complex, nonlinear relationships between ozone absorption features, surface and cloud properties, observation geometry, and atmospheric state variables.

The proposed methodology integrates TROPOMI spectral radiance/irradiance data (L1B) and the satellite position with meteorological information from ERA5 reanalysis data. The meteorological data includes boundary layer height and dissipation and surface pressure alongside temperature, humidity and wind speed profiles. The ground truth for the supervised training is comprised of co-located ozone profile measurements from ozone sondes (TOAR), aircraft measurements (IAGOS), lidar observations (TOLNet), and satellite microwave limb sounder (MLS).

Initial retrieval model is based on feed-forward fully connected neural network (multilayer perceptron), with planned extensions to convolutional architectures and dimensionality-reduction techniques. Using data from 2021 alone (approximately 1.2×10⁴ independent ozone profiles corresponding to ~1.5×10⁷ concentration measurements) our preliminary results demonstrate strong performance. The model’s ozone profile predictions are evaluated against the ground truth observations on an independent test set (including unseen time periods and locations). In this preliminary evaluation, the model achieves a coefficient of determination (R²) of 0.841 between retrieved and observed ozone concentrations, indicating the model’s ability to capture both vertical ozone structure and spatial ozone variability.

How to cite: Ginio‬‏, ‪., Wagner, T., Beirle, S., Kuhn, L., and Rudich, Y.: A Deep Learning Retrieval for Tropospheric Ozone Profiles from High-Resolution Satellite Data, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-11650, https://doi.org/10.5194/egusphere-egu26-11650, 2026.