Vegetation as proxies for improving the estimation of soil water fluxes

Soil water fluxes, including soil moisture, water storage, and recharge flux, are essential components of energy exchange at the Earth's surface and are fundamental to modeling land surface processes. Accurate estimation of soil hydraulic properties (SHPs) at the field scale is critical for simulating these fluxes, particularly within the vadose zone. Consequently, a robust understanding of soil water dynamics and associated processes relies on the precise characterization of SHPs. The experimental determination of these properties at different spatial scales are challenging and often time-consuming, especially in the case of vertically heterogeneous soils. Studies showed that the vegetation indices can provide sub-surface hydrological information. For example, the Leaf area index (LAI) of forest cover was found to be strongly correlated with the groundwater levels. This indicates that vegetation has the potential to act as a proxy for understanding many surface and sub-surface soil water processes. Inverse modeling approaches provide an opportunity to use vegetation information to estimate SHPs. The present study is aimed at developing and testing methodologies for estimating SHPs for multi-layered soils, specifically field capacity and wilting point, in an agricultural watershed. This is accomplished using variables like surface soil moisture, surface soil temperature, and canopy variables (Leaf Area Index and evapotranspiration) as proxies in different weighted likelihood combinations and carrying out the inverse modeling using the soil water balance model STICS. The methodology has been developed for three layered soil profiles (0 to 10 cm, 11 to 50 cm, and 51 to 100 cm) with combinations made from four major soil textures: sandy loam, sandy clay loam, clay loam, and clay, making 12 soil combinations. A sensitivity analysis of canopy variables relative to soil water storage properties was carried out to determine the best choice of canopy variable for estimating soil water fluxes using the EASI Method.  The results show that the soil moisture and canopy variables showed a strong correlation with SHPs, indicating that these variables could provide reliable estimates of soil water fluxes. In which the leaf area index shows more sensitivity towards the subsurface layers (sensitivity index~0.4). The study showed that the likelihood combinations of variables with higher weights to canopy variables provided better estimates of SHPs in the deeper layers. With the use of the likelihood combinations made by surface and canopy variables, we achieved mean relative absolute errors of 4% for the surface layer properties and 10% for the root zone SHPs, especially in water-stressed conditions. Since the variables used in this study are potentially accessible from the remote sensing data, the application of this methodology at large spatial scales is feasible, thereby generating spatial maps of sub-surface soil properties at regional scales, which can aid in the improved modeling of sub-surface soil moisture.