Assessment of SMOS and SMAP soil moisture products against calibrated land surface model output over the Huai river basin

Authors: WANG, X. - Hohai University; LU, H. - Hohai University; Crow, Wade; ZHU, Y. - Hohai University; SU, J. - Hohai University

Submitted to: Journal of Hydrology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/16/2021
Publication Date: 7/1/2021
Citation: Wang, X., Lu, H., Crow, W.T., Zhu, Y., Su, J. 2021. Assessment of SMOS and SMAP soil moisture products against calibrated land surface model output over the Huai river basin. Journal of Hydrology. 598:126468. https://doi.org/10.1016/j.jhydrol.2021.126468.
DOI: https://doi.org/10.1016/j.jhydrol.2021.126468

Interpretive Summary: The ability to estimate soil moisture from satellite observations is valuable for a wide range of agricultural applications. However, before they can be used with confidence, satellite-based soil moisture products must first be validated against independent ground-based observations. Such comparisons are difficult in areas of the world where available ground-based soil moisture observations have poor spatial coverage. This paper solves this problem by applying a land surface model to spatially interpolate sparse ground-based soil moisture observations and evaluates new soil moisture products from the Soil Moisture Active/Passive (SMAP) and Soil Moisture Ocean Salinity (SMOS) missions in a spatially continuous manner. Evaluation results presented in the paper validate this new approach and identify important sources of error currently degrading SMAP and SMOS soil moisture products. This technique, and the insights it generates, may be applied to improve the quality of satellite-based soil moisture data products.

Technical Abstract: Soil moisture often governs the exchange of water between the land surface and the atmosphere and, as a result, has a profound effect on the global water and energy cycles. With the launch of the two new L-band satellite missions tasked with retrieving surface soil moisture (SSM) (i.e., the ESA Soil Moisture and Ocean Salinity (SMOS) and the NASA Soil Moisture Active/Passive (SMAP) missions), new possibilities exist for quasi-global soil moisture monitoring. Here, a land surface model soil moisture product (CLDAS), generated via Back Propagation Neural Network (BPNN)-based calibration against ground observations, is employed to evaluate four satellite-derived SSM products (SMOS-L3, SMOS-IC, SMAP_L3_SM_P, and SMAP_L3_SM_P_E). All four satellite SSM products are evaluated within a uniform framework (i.e., at 6:00 AM local solar time and 0.25°×0.25° spatial) using the CLDAS soil moisture estimates as a spatially continuous reference over the humid/semi-humid Huai River Basin in China. Results are shown as follows: (1) BPNN calibration enhances the absolute accuracy of land surface model soil moisture estimates obtained at out-of-sample ground sites withheld for validation and therefore provides a reliable source of spatially continuous information for the validation of satellite-based SSM products. (2) Compared to the resulting reference data set, all four satellite SSM products exhibit degraded accuracy in mountainous and wooded areas of the Huai River Basin relative to areas of flatter terrain. (3) Satellite retrievals present seasonal differences in metrics versus the reference data, correlation coefficients (versus the calibrated CLDAS reference) decrease in the summer season (June to August) compared to the winter season, whereas reduced bias is found in the summer season. (4) SMAP retrievals are recommended if the pixels were covered with both valid SMOS and SMAP retrievals. This study describes an alternative approach for the spatially continuous evaluation of satellite SSM products and thereby optimizes the application of SMOS and SMAP-derived SSM products within the humid and semi-humid regions worldwide.