Improved subsurface soil moisture prediction from surface soil moisture through the integration of the (de)coupling effect

Abstract

Soil moisture is a crucial component of the forest ecosystem because it primarily controls vegetation growth. However, the potential for predicting the subsurface from the surface soil moisture through (de)coupling in boreal larch forest (Larix gmenilii) in China has rarely been investigated. In the present study, the regression and residual as well as cross-correlation analyses were used to evaluate the relationship between the surface (5 cm) and subsurface (10, 20, and 40 cm as well as the profile soil moisture) soil moisture in the Larix gmenilii forest. The applicability of the exponential filter (ExpF) and cumulative distribution function matching (CDF) methods for estimating the subsurface using the surface soil moisture in the study area and the impacts of (de)coupling on the simulation accuracy were assessed. Our results shown that the coupling between surface and subsurface soil moisture decreased with soil depth increasing and decoupling is dominant in the range varying between 0.11 and 0.22 cm3cm−3. The CDF method exhibits superiority for estimating the subsurface soil moisture relative to the ExpF method in the study area. Furthermore, the surface soil moisture values in the coupling range capture the dynamics of the subsurface soil moisture, while decoupling significantly reduces the accuracy of simulating the subsurface soil moisture. Thus, we propose an approach involving the reconstruction of the coupling time series associated with the (de)coupling, which improves the accuracy of the subsurface soil moisture simulation, particularly at depth, where the coupling is weak. Thus, the (de)coupling effect should henceforth be incorporated in the future subsurface soil moisture simulation studies and is helpful to understand the hydrological process of soil, improve regional models and the associated parameters.