Improving Aerosol Retrieval From MISR With a Physics-Informed Deep Learning Method

• Published in: IEEE transactions on geoscience and remote sensing Vol. 62; pp. 1 - 11
• Main Authors: Man, Wenjing, Tao, Minghui, Xu, Lina, Xu, Xiaoguang, Jiang, Jianfang, Wang, Jun, Wang, Lunche, Wang, Yi, Fan, Meng, Chen, Liangfu
• Format: Journal Article
• Language: English
• Published: New York IEEE 2024; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Accuracy; Aerosol algorithm; Aerosol Robotic Network; Aerosols; Albedo; Algorithms; Atmospheric measurements; Atmospheric modeling; Coefficients; Correlation coefficient; Correlation coefficients; Deep learning; deep learning (DL); eastern China; Lookup tables; Multi-angle Imaging SpectroRadiometer (MISR); multiangle; Optical analysis; Optical scattering; Optical sensors; Optical thickness; Parameters; Physics; Radiative transfer; Regional development; Retrieval; Satellites; Scattering angle; Spectroradiometers; Training
• ISSN: 0196-2892; 1558-0644
• DOI: 10.1109/TGRS.2024.3376598

The Multi-angle Imaging SpectroRadiometer (MISR) measurement with a large range of scattering angles provides valuable information about aerosol microphysical properties. The current MISR algorithm utilizes predefined aerosol mixtures in lookup tables (LUTs) to infer aerosol types and microphysical parameters, which performs well globally but remains subject to considerable uncertainties in regional scales. To make efficient use of MISR measurement, we developed a physics-informed deep learning (PDL) method to retrieve aerosol optical/microphysical parameters over land in eastern China. By combining the physical constraint of radiative transfer (RT) simulation and modeling ability of DL methods, each aerosol parameter can be modeled with the whole used MISR measurements separately with high computational efficiency. PDL aerosol optical depth (AOD) and fine AOD (FAOD) have high correlation coefficients (<inline-formula> <tex-math notation="LaTeX">R > </tex-math></inline-formula> 0.95) with AErosol RObotic NETwork (AERONET) observations, with 89% and 81% values falling into expected error (EE) envelope of ± (0.05 + 20%AOD<inline-formula> <tex-math notation="LaTeX">_{\mathrm {AERONET}} </tex-math></inline-formula>), respectively. Despite only a slightly higher accuracy than recent MISR Version 23 products, PDL retrievals have solved the underestimation problem of AOD and FAOD at moderate-to-high values (<inline-formula> <tex-math notation="LaTeX">> </tex-math></inline-formula>0.4). Besides better constraint of abnormal values in coarse AOD (CAOD), the PDL algorithm significantly improves the retrieval accuracy of MISR single scattering albedo (SSA). With reliable and robust performance, the PDL algorithm provides a flexible and efficient aerosol retrieval framework for emerging multiangle polarimetric (MAP) measurements.