Measuring full-range soil hydraulic properties for the prediction of crop water availability using gamma-ray attenuation and inverse modeling

Abstract

Accurate knowledge of soil hydraulic properties (K-θ-h) for the entire range of crop available water is essential for the prediction of soil water movement and related processes by mechanistic models, including the partitioning of surface energy fluxes into transpiration and evaporation and the dynamics of root water uptake, mandatory processes for adjustments of crop water use efficiency. We implemented an experimental and numerical protocol to obtain K-θ-h of eleven soils with a broad spectrum of texture and land use. Measurements of the soil water content during evaporation experiments using gamma-ray beam attenuation, a non-invasive technique, were adopted as an alternative approach to conventional measurements of the soil water pressure head. Inverse parameter optimization was performed using Hydrus-1D. The optimized K-θ-h functions were interpreted with respect to crop available water, where results calculated by a proposed “dynamic” method were compared with those determined using the conventional “static” criteria with standardized pressure heads. The evaporation experiment protocol allowed the determination of the K-θ-h relationships by inverse modeling from near-saturation to the dry range (∼ −150 m) with satisfactory accuracy. Soil water retention curves of the fine-textured soils determined by the conventional method (pressure plates) deviated from those estimated by the inverse optimization near saturation and in the dry range, with the conventional method predicting larger water content values. In terms of crop available water, the “dynamic” method allowed incorporating system characteristics (atmospheric demand and crop properties) and K-θ-h in a process-based way, contrarily to the “static” method. Considering a specific scenario, for the fine-textured soils the “static” and “dynamic” approaches performed similarly, however, for the coarse-textured soils, they diverged significantly. No tendency could be revealed for crop water availability under different land uses, and, in general, crop available water for soils under forest use was very similar to their counterparts under agricultural use.