Using near-infrared spectroscopy to estimate soil water retention curves with the van Genuchten model

Abstract

Hydraulic properties of unsaturated soils are essential for understanding hydric functioning and solving flow and mass-transfer problems in the vadose zone. One of the best-known models for representing the experimental soil–water retention curve, which describes the matric potential (h) as a function of the water content (θ) of a soil horizon, is the van Genuchten (VG) model. It requires four parameters whose values vary by soil type: residual water content (θr), saturated water content (θs), the inverse of the air-entry pressure (α) and a shape parameter (n). The main objective of this study was to show the relevance of using near-infrared (NIR) spectroscopy to estimate the parameters of the VG model, based on the relation established between the soil water spectral index (SWSI) and θ (Soltani et al., 2019a). Based on this approach, the effective saturation of VG equals the effective SWSI. We applied the approach to 25 soil samples collected from topsoil and subsoil horizons in Brittany (western France), which exhibited high variability in texture and soil organic carbon content ranging from 0.07 % to 6.23 %. The results showed that i) the NIR-spectroscopy approach was relevant for estimating hydraulic parameters θs, α and n of the VG model and ii) the parameters obtained from a VG-like equation based on the relation between h and SWSI predicted values of θ of the soil–water retention curve that were similar to observed values, with a root-mean-square error of 0.031 and 0.045 cm3cm−3 for topsoil and subsoil horizons, respectively. The method was thus more accurate for topsoil horizons.