Spatio-temporal Evolution and Driving Mechanisms of Soil Moisture Content in the Dabie Mountain Area

Soil moisture, a critical interface connecting the atmosphere, hydrosphere, and biosphere, holds substantial significance for the administration of watershed water resources and the preservation of ecological security. Based on monthly soil moisture content (SMC) datasets spanning 1961 to 2020 across 0–10 cm, 10–40 cm, 40–100 cm, 100–200 cm, and 0-200 cm soil layers in the Dabie Mountain Area, our research systematically investigated the spatio-temporal variation characteristics and driving mechanisms of SMC by integrating Sen’s slope estimation, the Mann-Kendall test, and Geodetector methods. The results indicated that five SMC layers exhibited a fluctuating decreasing trend from 1961 to 2020, with abrupt change points concentrated in the 2000s. The decline was more pronounced in the 0-200 cm and 100–200 cm layers, with decreasing rates of 0.305 kg·m⁻²·a⁻¹ and 0.1395 kg·m⁻²·a⁻¹, respectively. SMC values were highest in summer and lowest in winter. However, the greater seasonal decrease in spring and autumn dominated the interannual declining trend of SMC. Abrupt change points were identified in five SMC layers at the seasonal scale, with the abrupt change characteristics in spring showing the highest consistency with the interannual variation pattern and occurring slightly earlier. SMC across the five soil depth intervals displayed a distinct spatial pattern characterized by higher values in the eastern region and lower values in the western region of the study area. As soil depth increased, the proportion of area with high SMC values exhibited an increasing trend, and the distribution range of these high-value areas gradually expanded. SMC exhibited a decreasing trend across 98% of the study area, with the reduction being more pronounced in the northwest than in the southeast. Temperature was the core factor regulating the spatial differentiation of SMC (q = 30.81% for the 0-200 cm layer), while precipitation exerted a more significant direct regulatory effect on the 0–10 cm layer (q = 19.86%). All multi-factor interactions exhibited either bilinear or nonlinear enhancement. The most pronounced nonlinear interactive effect was identified between temperature and the NDVI, with q-values increasing from 60.74% to 67.95% as soil depth increased. The second strongest interaction was between human footprint and NDVI, with q-values ranging from 58.88% to 62.00%. Our research provided a scientific basis for optimal water resource allocation, ecological protection and restoration, and the formulation of climate change adaptation strategies in the Dabie Mountain Area.