Effects of Spatially Distributed Soil Properties on Hydrological Modeling via the Noah‐MP Land Model

Abstract Soil hydraulic properties (SHPs) are crucial in modeling hydrological and ecological processes across scales. However, most land surface models (LSMs) use SHPs based on land cover and soil types, thus neglecting the inherent heterogeneity in SHPs. We hypothesize that spatially distributed SHPs can enhance the performance of LSMs in modeling ecohydrological processes. In this study, we performed regional simulations to evaluate the effects of 3‐dimensional spatially distributed SHPs (3D SHPs) on Noah‐MP‐simulated hydrological fluxes. The simulations were conducted at a 3‐hourly and 0.1° resolution from 1981 to 2018 over mainland China. Compared with the default lookup table soil parameters, the use of 3D SHPs enhanced not only the accuracy of the simulated runoff and evapotranspiration (ET) but also the ability of Noah‐MP to reliably characterize the spatial patterns and vertical interactions of soil moisture in various climate zones. Moreover, 3D SHPs increase the mean Kling‐Gupta efficiency (KGE) for runoff (from 0.32 to 0.67) and ET (from 0.65 to 0.74) and decrease the mean root mean square error (RMSE) (from 0.067 to 0.037) for surface soil moisture (SSM) in these climate zones. The KGE is better suited for assessing the fit of runoff and ET, and RMSE is more appropriate for capturing the bias in SSM. The improvements in KGE (runoff: from 0.40 to 0.74) and RMSE (SSM: from 0.051 to 0.019) are more significant in humid climate zones. This study highlights the importance of heterogeneous soil properties for the overall performance of Noah‐MP.