A first evaluation of satellite soil moisture products over the Central-Western Tibetan Plateau using rain gauge observations

Abstract

Soil moisture plays a key role in regulating water and energy cycle in the Tibetan Plateau, which further impacts regional climate. However, the quality of existing soil moisture products over the central to west TP (CWTP) remains unclear due to the lack of in-situ observations. Using rain gauge data from a recently established rainfall network in the CWTP region and by checking the hydrological consistency between surface soil moisture and precipitation, this study performed a first evaluation of several satellite soil moisture products, including those from the Soil Moisture Active Passive (SMAP), the European Space Agency Climate Change Initiative (ESA CCI), and the artificial neural network reproduced AMSR-E/2 retrievals (NNsm) in this area. Results show that: (1) both descending and ascending orbits of the SMAP product generally outperform ESA CCI and NNsm, with more robust hydrological consistency across all rain gauge stations; (2) ESA CCI is subjected to low availability of effective retrievals in this area, with the combined one performs slightly more robust than the active and passive channels; and (3) NNsm possesses the most effective soil moisture retrievals owing to less revisit time of AMSR2, but its hydrological consistency is the lowest as compared to the other two products. Further analysis suggests that large topography and dense vegetation can potentially impact the retrieval accuracy of surface soil moisture and thus hydrological consistency. Besides, relatively large amounts of rainfall is likely to impose more positive increments in surface soil moisture, whereas extremely heavy rainfall may further degrade hydrological consistency. These findings are complementary to existing soil moisture evaluations in the eastern TP, and are expected to contribute to improving soil moisture retrieval algorithms and understanding land–atmosphere interactions over the entire plateau.