Stochastic modeling of soil water dynamics within the 0–4 m profile along a climate-soil transect on the Chinese Loess Plateau

Study on the spatiotemporal dynamics of soil water will provide important information for the sustainable development and utilization of water resources, especially in the water deficient arid and semiarid areas. However, the soil water dynamics in deep soil layers for different climate and soil conditions are still not clearly studied. In this work, soil water data in the 0–4 m profiles were collected from different transects located on the Chinese Loess Plateau (CLP), and the soil water dynamics in deep soil layers for both shrub and forest lands were analyzed using the Laio stochastic model. We found that, with the decrease of latitudes from 38°47′N to 34°17′N on the CLP, soil water content (SWC) decreased from 0.32 in the south to 0.14 in the north and fluctuated obviously in the surface layer over time. For the SWC probability density functions (PDFs) on the CLP, the Laio stochastic model produced RMSEs of 0.666 (Changwu), 0.977 (Shenmu), and 1.178 (Ansai). The simulated distribution peaks were 0.32, 0.13, and 0.19, respectively, which equalled the observed peak values at the corresponding sites and its distribution was significantly affected by both climatic and soil factors, based on the shapes of PDFs and the statistical moments of SWC. Across six locations, the simulated PDFs provided a regional envelope for soil water content evaluation. Locations with SWC below the lower line indicate relatively dry conditions relative to the modeled local distribution. Locations with SWC lower than the lines were identified as having high water-stress risk for tree growth, suggesting that these areas may not support sustainable forest establishment without additional water management. However, this should not be interpreted as a direct afforestation threshold; site-specific assessments incorporating species traits and water-storage capacity are required before operational recommendations. The findings in our study will provide important scientific guidance for the sustainable vegetation restoration in water-limited regions like the Chinese Loess Plateau.