Investigating multiple causes of time-varying SMAP soil moisture biases based on core validation sites data

The Soil Moisture Active Passive (SMAP) mission provides state-of-the-art global soil moisture (SM) datasets. However, seasonal SM biases and their contributing factors have not be systematically reviewed. This study evaluated the biases of SMAP V6 dual channel algorithm (DCA), single channel algorithm H-pol (SCA-H) and V-pol (SCA-V) SM products based on core validation sites data. All algorithms perform better under clear- than cloudy-sky, and in cloudless daytime than nighttime. Consecutive clear-sky benefits SM retrieval, progressively lowering the uncertainty of SM retrievals while at the cost of dry biases. Cloudy-sky deteriorates the quality of SM retrievals, and wet biases increase with the duration time of cloudy-sky. The modified V7 DCA has a major improvement, owning the potential to provide accurate retrievals under cloudy-sky. SMAP SM biases are co-determined by vegetation index, soil temperature and their biases, and a single factor only explains at most 54% variance in SM biases. Generally, SMAP SM bias is negatively correlated with soil temperature and positively correlated with its bias. SM bias correlates positively with vegetation index and its bias for single channel algorithms, and the underestimation of SM increases with vegetation density for DCA. To get a complete picture of SMAP SM biases, a total differential of radiative transfer equation is recommended for decomposing SMAP SM biases. To this end, more validation sites are required covering diverse land cover types and providing continuous data records.