In situ observations of soil hydraulic properties and soil moisture in a high, cold mountainous area of the northeastern Qinghai-Tibet Plateau

Information on soil hydraulic properties (SHPs) and soil moisture (SM) is essential to understand and model water and energy cycles at terrestrial surfaces. However, information regarding these soil properties in existing datasets is often scarce and inaccurate for high, cold mountainous areas such as the Qinghai-Tibet Plateau (QTP). To help bridge this gap, we have compiled an SHP and SM dataset for the northeastern QTP (a major high, cold mountainous area) using measurements of soil collected at 5 and 25 cm depths from 206 sampling sites, and in-situ observations from 32 SM monitoring stations at 5, 15, 25, 40, and 60 cm depths. We used this dataset to explore large-scale variations (spatial and temporal) in SHPs and SM across the study area. We also evaluated several widely used SHP (soil texture, bulk density, and saturated hydraulic conductivity) and SM datasets derived by remote-sensing methods, reanalysis and data assimilation. Our datasets showed that SM significantly decreases from the southeastern part to the northwestern part in the study area, and SM decreases with increases in depth over 0–70 cm. Moreover, the regional annual SM showed decreased trend from 2014 to 2020 in the study area. Additionally, we detected substantial bias in the currently available large SHP datasets, which do not capture the spatial variability recorded in the in-situ observations. Especially, clay and sand estimates from both HWSD and SoilGrid datasets were significantly overestimated, and silt was significantly underestimated within the depth of 0–30 cm in the study area. We also found that SM values derived from remote sensing datasets fitted the in-situ SM observations better than those derived from the reanalysis data (which had higher bias) and data assimilation (which did not capture the temporal variability of SM). Our findings emphasize the unneglectable bias of the widely-used large-scale SHP datasets, especially for the soil texture data. Thus, an urgent need for large-scale field sampling of SHP in mountainous areas. The in-situ observation dataset presented here provides important information with unprecedented coverage and resolution regarding the SHP variability and long-term SM trends across a large, high, cold mountainous area, thereby enhancing our understanding of water cycles and energy exchange processes over the QTP.