Disturbance Mechanism and Correction Model of CRNP Sensing Depth Estimation under Soil Water Vertical Heterogeneity

Cosmic-Ray Neutron Probe (CRNP) can be used to monitor near-surface soil moisture under undisturbed conditions. However, its sensing depth is significantly affected by the vertical distribution of soil water content. In this study, one CRNP site was established in a temperate coniferous and broad-leaved mixed forest in Northeast China (44°25′N-44°35′N, 127°50′E-128°00′E). Meanwhile, TDR probes were installed at depths of 5, 15, 30, and 60 cm to continuously monitor soil moisture dynamics from May 1 to December 27, 2022, for a total of 240 days. The analysis showed that when the difference in volumetric water content between the surface layer (0-15 cm) and the deep layer (15-60 cm) exceeded 0.06 m³/m³, the CRNP sensing depth decreased from the theoretical value of 21.8 cm to 17.3 cm, resulting in an underestimation of up to 20.6%. A correction model based on the soil water content gradient was constructed using nonlinear least squares fitting. The root mean square error (RMSE) of the training set decreased from 0.025 m to 0.012 m. In cross-site validation at three different ecological stations (grassland, farmland, and wetland), the average coefficient of determination (R²) reached 0.84. This study reveals the regulating effect of vertical soil moisture heterogeneity on CRNP sensing depth and provides a refined correction approach for regional soil moisture inversion models.