Systematic Evaluation of Four Satellite AOD Datasets for Estimating PM<sub>2.5</sub> Using a Random Forest Approach
The latest epidemiological studies have revealed that the adverse health effects of PM<sub>2.5</sub> have impacts beyond respiratory and cardio-vascular diseases and also affect the development of the brain and metabolic diseases. The need for accurate and spatio-temporally resolved PM&l...
| Asıl Yazarlar: | , , |
|---|---|
| Materyal Türü: | Makale |
| Dil: | English |
| Baskı/Yayın Bilgisi: |
MDPI AG
2023-04-01
|
| Seri Bilgileri: | Remote Sensing |
| Konular: | |
| Online Erişim: | https://www.mdpi.com/2072-4292/15/8/2064 |
| _version_ | 1827743815057801216 |
|---|---|
| author | Jana Handschuh Thilo Erbertseder Frank Baier |
| author_facet | Jana Handschuh Thilo Erbertseder Frank Baier |
| author_sort | Jana Handschuh |
| collection | DOAJ |
| description | The latest epidemiological studies have revealed that the adverse health effects of PM<sub>2.5</sub> have impacts beyond respiratory and cardio-vascular diseases and also affect the development of the brain and metabolic diseases. The need for accurate and spatio-temporally resolved PM<sub>2.5</sub> data has thus been substantiated. While the selective information provided by station measurements is mostly insufficient for area-wide monitoring, satellite data have been increasingly applied to comprehensively monitor PM<sub>2.5</sub> distributions. Although the accuracy and reliability of satellite-based PM<sub>2.5</sub> estimations have increased, most studies still rely on a single sensor. However, several datasets have become available in the meantime, which raises the need for a systematic analysis. This study presents the first systematic evaluation of four satellite-based AOD datasets obtained from different sensors and retrieval methodologies to derive ground-level PM<sub>2.5</sub> concentrations. We apply a random forest approach and analyze the effect of the resolution and coverage of the satellite data and the impact of proxy data on the performance. We examine AOD data from the Moderate resolution Imaging spectroradiometer (MODIS) onboard Terra and Aqua satellites, including Dark Target (DT) algorithm products and the Multi-Angle Implementation of Atmospheric Correction (MAIAC) product. Additionally, we explore more recent datasets from the Sea and Land Surface Temperature Radiometer (SLSTR) onboard Sentinel-3a and from the Tropospheric Monitoring Instrument (TROPOMI) operating on the Sentinel-5 precursor (S5p). The method is demonstrated for Germany and the year 2018, where a dense in situ measurement network and relevant proxy data are available. Overall, the model performance is satisfactory for all four datasets with cross-validated R<sup>2</sup> values ranging from 0.68 to 0.77 and excellent for MODIS AOD reaching correlations of almost 0.9. We find a strong dependency of the model performance on the coverage and resolution of the AOD training data. Feature importance rankings show that AOD has less weight compared to proxy data for SLSTR and TROPOMI. |
| first_indexed | 2024-03-11T04:33:49Z |
| format | Article |
| id | doaj.art-3b0778aa643c4ff78c376cefb7dc5163 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-11T04:33:49Z |
| publishDate | 2023-04-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-3b0778aa643c4ff78c376cefb7dc51632023-11-17T21:11:31ZengMDPI AGRemote Sensing2072-42922023-04-01158206410.3390/rs15082064Systematic Evaluation of Four Satellite AOD Datasets for Estimating PM<sub>2.5</sub> Using a Random Forest ApproachJana Handschuh0Thilo Erbertseder1Frank Baier2German Aerospace Center (DLR), German Remote Sensing, Data Center (DFD), 82234 Weßling, GermanyGerman Aerospace Center (DLR), German Remote Sensing, Data Center (DFD), 82234 Weßling, GermanyGerman Aerospace Center (DLR), German Remote Sensing, Data Center (DFD), 82234 Weßling, GermanyThe latest epidemiological studies have revealed that the adverse health effects of PM<sub>2.5</sub> have impacts beyond respiratory and cardio-vascular diseases and also affect the development of the brain and metabolic diseases. The need for accurate and spatio-temporally resolved PM<sub>2.5</sub> data has thus been substantiated. While the selective information provided by station measurements is mostly insufficient for area-wide monitoring, satellite data have been increasingly applied to comprehensively monitor PM<sub>2.5</sub> distributions. Although the accuracy and reliability of satellite-based PM<sub>2.5</sub> estimations have increased, most studies still rely on a single sensor. However, several datasets have become available in the meantime, which raises the need for a systematic analysis. This study presents the first systematic evaluation of four satellite-based AOD datasets obtained from different sensors and retrieval methodologies to derive ground-level PM<sub>2.5</sub> concentrations. We apply a random forest approach and analyze the effect of the resolution and coverage of the satellite data and the impact of proxy data on the performance. We examine AOD data from the Moderate resolution Imaging spectroradiometer (MODIS) onboard Terra and Aqua satellites, including Dark Target (DT) algorithm products and the Multi-Angle Implementation of Atmospheric Correction (MAIAC) product. Additionally, we explore more recent datasets from the Sea and Land Surface Temperature Radiometer (SLSTR) onboard Sentinel-3a and from the Tropospheric Monitoring Instrument (TROPOMI) operating on the Sentinel-5 precursor (S5p). The method is demonstrated for Germany and the year 2018, where a dense in situ measurement network and relevant proxy data are available. Overall, the model performance is satisfactory for all four datasets with cross-validated R<sup>2</sup> values ranging from 0.68 to 0.77 and excellent for MODIS AOD reaching correlations of almost 0.9. We find a strong dependency of the model performance on the coverage and resolution of the AOD training data. Feature importance rankings show that AOD has less weight compared to proxy data for SLSTR and TROPOMI.https://www.mdpi.com/2072-4292/15/8/2064satellite AODPM<sub>2.5</sub>random forestfeature importanceGermany |
| spellingShingle | Jana Handschuh Thilo Erbertseder Frank Baier Systematic Evaluation of Four Satellite AOD Datasets for Estimating PM<sub>2.5</sub> Using a Random Forest Approach Remote Sensing satellite AOD PM<sub>2.5</sub> random forest feature importance Germany |
| title | Systematic Evaluation of Four Satellite AOD Datasets for Estimating PM<sub>2.5</sub> Using a Random Forest Approach |
| title_full | Systematic Evaluation of Four Satellite AOD Datasets for Estimating PM<sub>2.5</sub> Using a Random Forest Approach |
| title_fullStr | Systematic Evaluation of Four Satellite AOD Datasets for Estimating PM<sub>2.5</sub> Using a Random Forest Approach |
| title_full_unstemmed | Systematic Evaluation of Four Satellite AOD Datasets for Estimating PM<sub>2.5</sub> Using a Random Forest Approach |
| title_short | Systematic Evaluation of Four Satellite AOD Datasets for Estimating PM<sub>2.5</sub> Using a Random Forest Approach |
| title_sort | systematic evaluation of four satellite aod datasets for estimating pm sub 2 5 sub using a random forest approach |
| topic | satellite AOD PM<sub>2.5</sub> random forest feature importance Germany |
| url | https://www.mdpi.com/2072-4292/15/8/2064 |
| work_keys_str_mv | AT janahandschuh systematicevaluationoffoursatelliteaoddatasetsforestimatingpmsub25subusingarandomforestapproach AT thiloerbertseder systematicevaluationoffoursatelliteaoddatasetsforestimatingpmsub25subusingarandomforestapproach AT frankbaier systematicevaluationoffoursatelliteaoddatasetsforestimatingpmsub25subusingarandomforestapproach |