SSAT: Sensor-Satellite Auto-Correlation Transformer for Enhanced Aerosol Optical Depth Prediction
Satellite-derived aerosol optical depth (AOD) observations are highly valuable to describe the horizontal distribution of aerosols, but are hampered by spatial data gaps and limited temporal coverage (typically once a day). The synergism of these measurements with chemistry-transport models (CTM) ma...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11091309/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1839611129678004224 |
|---|---|
| author | Maoquan Zhang Bisser Raytchev Juan Cuesta Farouk Lemmouchi Maithili Karle Daniel Andrade |
| author_facet | Maoquan Zhang Bisser Raytchev Juan Cuesta Farouk Lemmouchi Maithili Karle Daniel Andrade |
| author_sort | Maoquan Zhang |
| collection | DOAJ |
| description | Satellite-derived aerosol optical depth (AOD) observations are highly valuable to describe the horizontal distribution of aerosols, but are hampered by spatial data gaps and limited temporal coverage (typically once a day). The synergism of these measurements with chemistry-transport models (CTM) may be used to overcome these limitations. Although physically constrained methods (e.g. data assimilation) are a common practice for addressing this synergism, these methods may be difficult to implement and are highly computationally demanding. On the other hand, statistical or learning-based techniques can still offer flexible and effective solutions. In this work, we present SSAT, a transformer-based approach that fuses satellite and model outputs to refine AOD predictions without modifying the underlying CTM itself. Our design leverages the auto-correlation mechanism from Autoformer and introduces GridMSE, a specialized loss function aimed at improving spatial coherence and handling imbalanced data. Extensive experiments show that SSAT has higher accuracy and better visual correspondence compared to existing tree-based and deep learning baselines, particularly in underrepresented aerosol load regimes. While it does not replace physically rigorous data assimilation, SSAT can serve as a complementary tool for quickly refining AOD maps, ultimately offering a more complete basis for monitoring aerosol distributions and informing environmental analyses. |
| format | Article |
| id | doaj-art-d7d30c8c22474a23a97024dce2d5bcf1 |
| institution | Matheson Library |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-d7d30c8c22474a23a97024dce2d5bcf12025-07-28T23:00:47ZengIEEEIEEE Access2169-35362025-01-011313083213084510.1109/ACCESS.2025.359147911091309SSAT: Sensor-Satellite Auto-Correlation Transformer for Enhanced Aerosol Optical Depth PredictionMaoquan Zhang0https://orcid.org/0009-0001-4356-3503Bisser Raytchev1https://orcid.org/0000-0002-2146-415XJuan Cuesta2https://orcid.org/0000-0001-9330-6401Farouk Lemmouchi3https://orcid.org/0009-0006-1958-7499Maithili Karle4https://orcid.org/0009-0006-2792-8230Daniel Andrade5https://orcid.org/0000-0002-1123-4369Graduate School of Advanced Science and Engineering, Hiroshima University, Hiroshima, JapanGraduate School of Advanced Science and Engineering, Hiroshima University, Hiroshima, JapanCNRS, LISA, Université Paris-Est Créteil and Université Paris Cité, Créteil, FranceCNRS, LISA, Université Paris-Est Créteil and Université Paris Cité, Créteil, FranceDepartment of Computer Engineering, MKSSS’s Cummins College of Engineering for Women, Pune, Maharashtra, IndiaGraduate School of Advanced Science and Engineering, Hiroshima University, Hiroshima, JapanSatellite-derived aerosol optical depth (AOD) observations are highly valuable to describe the horizontal distribution of aerosols, but are hampered by spatial data gaps and limited temporal coverage (typically once a day). The synergism of these measurements with chemistry-transport models (CTM) may be used to overcome these limitations. Although physically constrained methods (e.g. data assimilation) are a common practice for addressing this synergism, these methods may be difficult to implement and are highly computationally demanding. On the other hand, statistical or learning-based techniques can still offer flexible and effective solutions. In this work, we present SSAT, a transformer-based approach that fuses satellite and model outputs to refine AOD predictions without modifying the underlying CTM itself. Our design leverages the auto-correlation mechanism from Autoformer and introduces GridMSE, a specialized loss function aimed at improving spatial coherence and handling imbalanced data. Extensive experiments show that SSAT has higher accuracy and better visual correspondence compared to existing tree-based and deep learning baselines, particularly in underrepresented aerosol load regimes. While it does not replace physically rigorous data assimilation, SSAT can serve as a complementary tool for quickly refining AOD maps, ultimately offering a more complete basis for monitoring aerosol distributions and informing environmental analyses.https://ieeexplore.ieee.org/document/11091309/Aerosol optical depth (AOD) predictionsatellite remote sensingtransformer regression predictiondeep learning complementary for physical data assimilationenhanced spatio-temporal prediction by GridMSE loss function |
| spellingShingle | Maoquan Zhang Bisser Raytchev Juan Cuesta Farouk Lemmouchi Maithili Karle Daniel Andrade SSAT: Sensor-Satellite Auto-Correlation Transformer for Enhanced Aerosol Optical Depth Prediction IEEE Access Aerosol optical depth (AOD) prediction satellite remote sensing transformer regression prediction deep learning complementary for physical data assimilation enhanced spatio-temporal prediction by GridMSE loss function |
| title | SSAT: Sensor-Satellite Auto-Correlation Transformer for Enhanced Aerosol Optical Depth Prediction |
| title_full | SSAT: Sensor-Satellite Auto-Correlation Transformer for Enhanced Aerosol Optical Depth Prediction |
| title_fullStr | SSAT: Sensor-Satellite Auto-Correlation Transformer for Enhanced Aerosol Optical Depth Prediction |
| title_full_unstemmed | SSAT: Sensor-Satellite Auto-Correlation Transformer for Enhanced Aerosol Optical Depth Prediction |
| title_short | SSAT: Sensor-Satellite Auto-Correlation Transformer for Enhanced Aerosol Optical Depth Prediction |
| title_sort | ssat sensor satellite auto correlation transformer for enhanced aerosol optical depth prediction |
| topic | Aerosol optical depth (AOD) prediction satellite remote sensing transformer regression prediction deep learning complementary for physical data assimilation enhanced spatio-temporal prediction by GridMSE loss function |
| url | https://ieeexplore.ieee.org/document/11091309/ |
| work_keys_str_mv | AT maoquanzhang ssatsensorsatelliteautocorrelationtransformerforenhancedaerosolopticaldepthprediction AT bisserraytchev ssatsensorsatelliteautocorrelationtransformerforenhancedaerosolopticaldepthprediction AT juancuesta ssatsensorsatelliteautocorrelationtransformerforenhancedaerosolopticaldepthprediction AT farouklemmouchi ssatsensorsatelliteautocorrelationtransformerforenhancedaerosolopticaldepthprediction AT maithilikarle ssatsensorsatelliteautocorrelationtransformerforenhancedaerosolopticaldepthprediction AT danielandrade ssatsensorsatelliteautocorrelationtransformerforenhancedaerosolopticaldepthprediction |