A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Performance comparison of Fengyun-4A and Himawari-8 in PM2.5 estimation in China
https://orcid.org/0000-0002-3012-2493
Abstract The increase in the PM2.5 concentration has become the focus of public concern due to the health risks it poses. In contrast to the sparsely distributed ground stations, geostationary satellites can provide a unique method of frequently monitoring national-scale PM2.5 concentrations. The Himawari-8 and Fengyun-4A (FY-4A) satellites have both been used to monitor PM2.5 concentrations due to their high spatial (5 km and 4 km, respectively) and temporal resolutions (hourly), but it is still not clear which satellite is more suitable for such studies. In this study, the hourly PM2.5 concentrations were estimated based on a deep belief network using top-of-atmosphere (TOA) reflectance data and other auxiliary variables from April 2018 to March 2019. To comprehensively and effectively verify the accuracy of the results, three cross-validation methods were applied. The evaluation indexes of FY-4A were slightly higher than those of Himawari-8. The distributions of their estimated PM2.5 concentration mappings were consistent in terms of the overall trend, but they exhibit regional inconsistencies. In-depth analysis of the results revealed that the main reasons for these inconsistencies are the differences in the spatial resolution, training data, satellite platforms, and retrieval algorithms. Through comparative analysis, the effects of the satellite platforms and retrieval algorithms were determined to be negligible. In general, the FY-4A satellite was found to be more adaptable to PM2.5 concentration estimation in China due to its wide coverage and high estimation accuracy.
Graphical abstract Display Omitted
Highlights Compared with the Himawari-8 satellite, Fengyun-4A satellite is more suitable for estimating PM2.5 concentrations in China. Three different ten-fold cross-validation approaches are introduced for a comprehensive evaluation of two satellites. The spatial resolution and training data bring effects on the mappings, the effect of the platforms and retrieval algorithms can be ignored.
Performance comparison of Fengyun-4A and Himawari-8 in PM2.5 estimation in China
https://orcid.org/0000-0002-3012-2493
Abstract The increase in the PM2.5 concentration has become the focus of public concern due to the health risks it poses. In contrast to the sparsely distributed ground stations, geostationary satellites can provide a unique method of frequently monitoring national-scale PM2.5 concentrations. The Himawari-8 and Fengyun-4A (FY-4A) satellites have both been used to monitor PM2.5 concentrations due to their high spatial (5 km and 4 km, respectively) and temporal resolutions (hourly), but it is still not clear which satellite is more suitable for such studies. In this study, the hourly PM2.5 concentrations were estimated based on a deep belief network using top-of-atmosphere (TOA) reflectance data and other auxiliary variables from April 2018 to March 2019. To comprehensively and effectively verify the accuracy of the results, three cross-validation methods were applied. The evaluation indexes of FY-4A were slightly higher than those of Himawari-8. The distributions of their estimated PM2.5 concentration mappings were consistent in terms of the overall trend, but they exhibit regional inconsistencies. In-depth analysis of the results revealed that the main reasons for these inconsistencies are the differences in the spatial resolution, training data, satellite platforms, and retrieval algorithms. Through comparative analysis, the effects of the satellite platforms and retrieval algorithms were determined to be negligible. In general, the FY-4A satellite was found to be more adaptable to PM2.5 concentration estimation in China due to its wide coverage and high estimation accuracy.
Graphical abstract Display Omitted
Highlights Compared with the Himawari-8 satellite, Fengyun-4A satellite is more suitable for estimating PM2.5 concentrations in China. Three different ten-fold cross-validation approaches are introduced for a comprehensive evaluation of two satellites. The spatial resolution and training data bring effects on the mappings, the effect of the platforms and retrieval algorithms can be ignored.
Performance comparison of Fengyun-4A and Himawari-8 in PM2.5 estimation in China
https://orcid.org/0000-0002-3012-2493
Abstract The increase in the PM2.5 concentration has become the focus of public concern due to the health risks it poses. In contrast to the sparsely distributed ground stations, geostationary satellites can provide a unique method of frequently monitoring national-scale PM2.5 concentrations. The Himawari-8 and Fengyun-4A (FY-4A) satellites have both been used to monitor PM2.5 concentrations due to their high spatial (5 km and 4 km, respectively) and temporal resolutions (hourly), but it is still not clear which satellite is more suitable for such studies. In this study, the hourly PM2.5 concentrations were estimated based on a deep belief network using top-of-atmosphere (TOA) reflectance data and other auxiliary variables from April 2018 to March 2019. To comprehensively and effectively verify the accuracy of the results, three cross-validation methods were applied. The evaluation indexes of FY-4A were slightly higher than those of Himawari-8. The distributions of their estimated PM2.5 concentration mappings were consistent in terms of the overall trend, but they exhibit regional inconsistencies. In-depth analysis of the results revealed that the main reasons for these inconsistencies are the differences in the spatial resolution, training data, satellite platforms, and retrieval algorithms. Through comparative analysis, the effects of the satellite platforms and retrieval algorithms were determined to be negligible. In general, the FY-4A satellite was found to be more adaptable to PM2.5 concentration estimation in China due to its wide coverage and high estimation accuracy.
Graphical abstract Display Omitted
Highlights Compared with the Himawari-8 satellite, Fengyun-4A satellite is more suitable for estimating PM2.5 concentrations in China. Three different ten-fold cross-validation approaches are introduced for a comprehensive evaluation of two satellites. The spatial resolution and training data bring effects on the mappings, the effect of the platforms and retrieval algorithms can be ignored.
Performance comparison of Fengyun-4A and Himawari-8 in PM2.5 estimation in China
Hu, Yuqing (author) / Zeng, Chao (author) / Li, Tongwen (author) / Shen, Huanfeng (author)
Atmospheric Environment ; 271
2021-12-10
Article (Journal)
Electronic Resource
English
Estimation and Correction of Geolocation Errors in FengYun-3C Microwave Radiation Imager Data
| Online Contents | 2016
Postlaunch calibration of FengYun-3B MERSI reflective solar bands
| Online Contents | 2013
First assessment of surface solar irradiance derived from Himawari-8 across China
| British Library Online Contents | 2018
Postlaunch Calibration of FengYun-3B MERSI Reflective Solar Bands
| Online Contents | 2013
First assessment of surface solar irradiance derived from Himawari-8 across China
| British Library Online Contents | 2018