A multi-temporal and multi-angular approach for systematically retrieving soil moisture and vegetation optical depth from SMOS data
Yu Bai,Tianjie Zhao,Li Jia,Michael H. Cosh,Jiancheng Shi,Zhiqing Peng,Xiaojun Li,Jean-Pierre Wigneron
DOI: https://doi.org/10.1016/j.rse.2022.113190
IF: 13.5
2022-10-01
Remote Sensing of Environment
Abstract:Passive microwave remote sensing of soil moisture is an underdetermined problem, as observed microwave emission from the landscape is affected by a variety of unknown surface parameters. Increasing observation information is an effective means to make retrievals more robust. In this study, a multi-temporal and multi-angular (MTMA) approach is proposed using SMOS (Soil Moisture and Ocean Salinity) satellite L-band data for retrieving vegetation optical depth (VODp, p indicates the polarization with H for horizontal and V for vertical)), effective scattering albedo (ωpeff), soil surface roughness (Zps), and soil moisture (SMp). The advantage of the MTMA approach is that it does not need auxiliary data as inputs or constraints. SMOS polarization-dependent VOD are produced and compared at a global scale for the first time, and it is found that the polarization difference of vegetation effects should not be ignored in the SM retrieval algorithm. The MTMA VOD retrievals are found to have a reasonable global spatial distribution, which is generally consistent with the VOD retrievals obtained from the SMOS Level 3 (SMOS-L3) and SMOS-IC Version 2 (V2) (referred to as SMOS-IC), except for showing relatively lower values over densely vegetated areas compared with the other two SMOS products. The spatial distribution of retrieved ωpeff generally shows a dependence on both VOD and land cover types. In addition, the values of MTMA-ωVeff are higher than that of MTMA-ωHeff, indicating stronger microwave scattering of V-pol in the vegetation layer than that of H-pol. The retrieved surface roughness parameter (Zps) ranges from 0.04 to 0.22 cm, and its spatial distribution is partially different from the existing roughness products/auxiliary data from SMOS and SMAP. The retrieved MTMA SM shows generally high correlations with in-situ measurements (11 dense observation networks) with overall correlation coefficients of > 0.75. The overall ubRMSE of MTMA-SMH and MTMA-SMV are < 0.055 m3/m3 and lower than that of SMOS-IC and SMOS-L3 products. SMOS-IC generally presents higher correlation coefficients compared to MTMA in most sites outside China; in China, RFI filtering is crucial and makes it very difficult when comparing algorithms based on different brightness temperature products. The number of effective retrievals of MTMA-SMH and MTMA-SMV ranges from 1409 to 1640 and 1104 to 1603 respectively, which is more than that from SMOS-IC (from 236 to 1358) over the selected 11 networks. Therefore, it is concluded that by incorporating multi-temporal SMOS data, the proposed method of MTMA can be used to systematically retrieve SM, VOD and additional surface parameters (effective scattering albedo and surface roughness) with comparable or better performance of SM than that of SMOS-IC and SMOS-L3. Moreover, this paper for the first time produced a polarization-dependent SMOS VOD product at a global scale.
environmental sciences,imaging science & photographic technology,remote sensing