Multi-Scale and Multi-Depth Validation of Soil Moisture From the China Land Data Assimilation System

Soil moisture (SM) plays a critical role in vegetation growth, terrestrial hydrological cycle, and regional climate change. In this article, the China Meteorological Administration Land Data Assimilation System (CLDAS) SM was validated at multiscale and multidepth against in situ measurements over the Tibetan plateau, and the Global Land Data Assimilation System (GLDAS) SM was also used as a reference for the validation process. The results revealed the following: First, in validations using the large, medium, and small scale networks, the small network achieved comparatively superior accuracy with a higher fitting degree and smaller statistical error, proving the reliability and robustness of densely clustered in situ stations. Second, the CLDAS SM can correctly present the promotion effect from precipitation. The most significant response of the CLDAS SM to precipitation lagged from 1 to 3 days in terms of growing depth, whereas promotion to SM tended to decrease as the soil depth increased. Third, CLDAS and GLDAS reasonably fitted in situ measurements in the frozen and unfrozen seasons, while their bias gradually increased with increasing depth. The SM errors in the snow-covered region were relatively stable. Diurnal land surface temperature exhibited a positive correlation with the bias of top-layer SM. In summary, this article shows that CLDAS SM retrievals favorably reproduced values and dynamics of in situ measurements and can be expected to act as a valuable reference for water cycle studies and climate evolution analysis.