ABSTRACT Understanding rainfall‐induced soil water dynamics (rainfall‐induced SWD) is crucial for improving Earth System Models (ESMs), which have been limited in large‐scale analyses. Using hourly data from 594 in situ stations, this study characterises the profile distribution patterns of rainfall‐induced SWD, examines the environmental impacts and their scale effects and analyses the occurrence and controlling factors of preferential flow (PF). In the soil profile, three distinct patterns of soil water response to rainfall events were identified: a predominant shallow‐decline pattern, a uniform pattern in barren land and a multi‐phase pattern in areas with shallow groundwater, influenced by interactions between soil water and groundwater. The influence of rainfall and clay on soil water response amplitude decreases, while antecedent soil water content (SWC) and soil organic carbon (SOC) have increasing effects as the spatial scale decreases. Notably, these changes manifest as significant, stepwise shifts at critical scales, which should be considered in future investigations. The normalised difference vegetation index (NDVI) affects the maximum rate of the soil‐wetting curve (Smax) in a bimodal fashion, with stronger impacts observed at regional scales and below 90 × 90 km 2 . Two distinct PF frequency distribution patterns emerge across soil layers: a consistent decrease with depth in PF frequency in the Changjiang river plain (CJ) region, corresponding with decreasing root density and microporosity. The predominant pattern is a mid‐depth increase, caused by uneven grass root distribution, along with contributions from gravel content, cracks and soil fauna in deeper soils. Soil texture exhibits complex, nonlinear effects on PF frequency, varying with antecedent SWC in areas dominated by fine texture. Bulk density has a threshold effect: it negatively impacts PF frequency below 1.33 g/cm 3 and positively impacts it above 1.41 g/cm 3 in mainland China. Our results reveal novel profile patterns of rainfall‐induced SWD and PF, providing fresh insights into the soil wetting process. This understanding is beneficial for effective model parameterisation and validation, ultimately enhancing the simulation of soil water dynamics in ESMs.
Zhang et al. (Wed,) studied this question.