Characteristics and influencing factors of soil moisture memory across mainland China

Abstract

Soil moisture memory (SMM), which is defined as the time required to “forget” a perturbation and reflects the strength of land–atmosphere coupling, plays a crucial role in understanding hydrological and eco-meteorological processes. However, large-scale estimates of SMM based on in situ observations remain scarce, and recently proposed SMM metrics have not yet been comprehensively evaluated using extensive ground measurements. In this study, a total of 2,218 available daily in situ soil moisture observations were collected across mainland China. An exponential drydown model was constructed based on these data to simulate the soil moisture drydown process and to estimate three key parameters including the e-folding time scale ($\tau$), the estimated lower bound of soil moisture (${\hat{\theta }}_w$) and magnitude of the drydown event ($\Delta \theta$). Independent site-specific wilting point observations are employed to evaluate the model performance, showing good agreement with fitted ${\hat{\theta }}_w$ values (adjusted R² = 0.58), which indicates that the model reliably captures variations in soil moisture status. Then we investigate the spatial characteristics of SMM, and explored the effects of soil depth, texture, and climatic conditions. Results show that all fitted parameters exhibit pronounced spatial heterogeneity across mainland China, especially between North and South China and between coastal and inland areas. As soil depth increases, $\Delta \theta$ decreases, ${\hat{\theta }}_w$ increases, and $\tau$ shows a non-monotonic pattern, first increases and then decreases. Among the controlling factors, ${\hat{\theta }}_w$ primarily governed by soil texture, especially clay content, whereas τ is more sensitive to meteorological factors, particularly evapotranspiration. Furthermore, we compared $\tau$ derived from in situ observations with that from multiple satellite and reanalysis soil moisture products. Most products systematically underestimate $\tau$ and show spatially inconsistent patterns, reflecting limitations in current land surface models in capturing soil drying processes. This study provides the first comprehensive assessment of SMM over mainland China based on a large amount of in situ soil moisture observations, revealing its spatial and depth-dependent characteristics and key controlling factors. The results offer new insights into the mechanisms governing SMM and have implications for improving drought diagnosis and land–atmosphere interaction modelling.