[A-CG37] 衛星による地球環境観測

[ACG37-P02] 多波長多ピクセルの衛星データを同時に用いたエアロゾル推定アルゴリズムの更新

*関口 美保¹、橋本 真喜子²、中島 映至¹ (1.国立大学法人東京海洋大学、2.国立研究開発法人宇宙航空研究開発機構)

キーワード：大気エアロゾル、衛星観測、アルゴリズム開発

We have developed the remote sensing algorithm to retrieve atmospheric aerosol properties using satellite observed radiances in several bands over a group of pixels with inter-pixel constraint. This method, named MWPM (Multi-Wavelength and multi-Pixel Method) is enabled to get surface reflectance and aerosol properties over land simultaneously and adopted as the standard aerosol algorithm over land for GOSAT-2/CAI-2. Although MWPM has successfully provided valuable atmospheric aerosol properties from CAI-2 observations, several shortcomings were pointed out in aerosol products. In this study, we introduce the update of MWPM. First, the size distribution of aerosols assumed in this method is update to that estimated from in-situ observations. Second, due to the tendency of single scattering albedo to underestimate, a part of black carbon assumed in this method was replaced to brown carbon to reduce overestimated absorption. At last, we modified the technique to select dark or bright pixels.
This updated MWPM was used for GCOM/SGLI and GOSAT-2/CAI-2 data to retrieval atmospheric aerosols at 15 SKYNET sites and 35 AERONET sites during March-December 2019 and validated with analytical values from in-situ observations. According to the results, the accuracy of this updated method was better than the previous one.