Analysis of the spatiotemporal characteristics of soil moisture memory and the attribution of changes in freezing depth in the Lhasa River Basin

Soil moisture is a key factor connecting climate signals and freezing processes, and its dynamic changes profoundly affect the water cycle and ecosystems in alpine regions. However, the vertical variation patterns of its soil moisture memory (SMM) and the mechanisms influencing freezing depth are still unclear. Taking the Lhasa River Basin as the study area, based on multi-source data such as GLDAS soil moisture (0~10, 10~40, 40~100, 100~200 cm), MODIS land surface temperature, precipitation, NDVI, and snow depth, this study comprehensively utilized methods such as Sen's slope trend analysis, Mann-Kendall test, Hurst index, partial correlation analysis, and random forest to systematically analyze the spatiotemporal variation patterns of soil moisture and freezing depth from 2000 to 2020. It quantified the spatial patterns of soil moisture memory at different depths and identified the dominant factors driving the variation of freezing depth. The results showed that soil moisture changes in the Lhasa River Basin exhibited significant vertical differentiation, gradually shifting from a large-scale decrease in the surface layer (92.2%) to a pattern of "increasing in the east and decreasing in the west" in the middle layer, with stable increasing and decreasing trends in the deep layer. Soil moisture memory was relatively short in the surface layer of the basin (1-2 months), gradually increasing with depth in most parts of the basin, and forming a stable memory core in the farmland area (SMM in June-July). NDVI contributed the most to the spatial variation of freezing depth (54%), followed by land surface temperature (contribution rate of 26%), while the impact of snow depth and precipitation was limited. Considering the role of temperature, there was a complex spatial pattern of negative partial correlation and local positive partial correlation between vegetation coverage and freezing depth. This study reveals the spatiotemporal distribution characteristics of soil moisture memory in the Lhasa River Basin and the mechanisms by which related factors affect freezing depth, which is of great significance for understanding land-atmosphere interactions and water resource sustainability in alpine regions.