当前位置：
X-MOL 学术
›
Remote Sens. Environ.
›
论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
An improved algorithm for retrieving high resolution fine-mode aerosol based on polarized satellite data: Application and validation for POLDER-3
Remote Sensing of Environment ( IF 13.5 ) Pub Date : 2020-09-01 , DOI: 10.1016/j.rse.2020.111894 Bangyu Ge , Xiaodong Mei , Zhengqiang Li , Weizhen Hou , Yisong Xie , Yang Zhang , Hua Xu , Kaitao Li , Yuanyuan Wei
Remote Sensing of Environment ( IF 13.5 ) Pub Date : 2020-09-01 , DOI: 10.1016/j.rse.2020.111894 Bangyu Ge , Xiaodong Mei , Zhengqiang Li , Weizhen Hou , Yisong Xie , Yang Zhang , Hua Xu , Kaitao Li , Yuanyuan Wei
Abstract The fine-mode aerosols (FMA), generally generated from anthropogenic sources, play an essential role in global radiation balance, climate change, atmospheric environment, and human health. However, the FMA retrieval remains a challenge. An improved high resolution FMA retrieval algorithm based on the Spectral Neutrality of Surface Polarized Reflectance (SNOSPR) was developed using polarized satellite data. SNOSPR utilizes a lookup table approach to simultaneously retrieve high spatial resolution FMA parameters, such as FMA optical depth (AODf), Angstrom exponent (AE), and surface polarized reflectance (SPR), without using the surface Bidirectional Polarization Distribution Function (BPDF) model, which may yield relatively more significant errors of SPR estimation in certain surface types. Besides, a new cost function (CF) based on the hypothesis of invariance of SPR was used to select the best retrievals. SNOSPR algorithm was applied to POLDER-3 polarized observation data (i.e., 2010–2012). Inter-comparison between SNOSPR (6 km) and POLDER-3 (L2_LGC,18 km) for two typical cases and monthly averaged AODf products showed a generally consistent spatial distribution, while SNOSPR provided more details, and better coverage over barren and coastal areas at high spatial resolution. Furthermore, all retrieved AODf, AE products of SNOSPR and L2_LGC were validated using AERosol RObotic NETwork (AERONET) measurements. The AODf results showed fine correlation coefficients, i.e., R of 0.9 and 0.89, respectively. However, products of SNOSPR had the expected error range which was higher than that of L2_LGC over all surface types (i.e., 77.3% vs 70.6%), urban (i.e., 76.0% vs 60.6%) and wetland (i.e., 82.4% vs 68.4%). Time-series comparison over grassland showed that SNOSPR had better coverage and more accurate retrievals. Considering AE, both SNOSPR and L2_LGC underestimated its value when validated against AERONET measurements. On the other hand, SPR was retrieved at higher retrieval accuracy by SNOSPR than L2_LGC, which used BPDF. Accuracies of the retrieved AODf improved significantly when multi-directional data was used, especially when the viewing angle number increased from one to four. The uncertainty analysis of spectral neutrality and spatial invariance of SPR was discussed for future improvements of SNOSPR. Last, SNOSPR was applied on Directional Polarimetric Camera (DPC), i.e., a POLDER-like sensor onboard the Chinese Gao Fen-5 satellite. The results in this study demonstrated the effectiveness of the SNOSPR algorithm and its potential for more reliable AODf/SPR retrievals for monitoring local atmospheric pollution and related studies.
更新日期：2020-09-01
京公网安备 11010802027423号