Estimating Soil Hydraulic Properties of the Soil-Plant-Root Zone using time-lapse Horizontal Borehole Ground Penetrating Radar Data in a Sequential Hydrogeophysical Inversion Approach

Abstract:

The soil hydraulic parameters play a vital role in sustainable crop production. They are the governing factors for the spatialtemporal water distribution in the soil-plant-root-zone and therefore regulate the water availability for crops. Estimating these parameters, using standard invasive techniques is time-consuming and often only provides point-scale information. To non-invasively derive them at a field-plot scale, we combined the soil water content (SWC) information from a hydraulic model with the SWC distribution from time-lapse horizontal crosshole ground penetrating radar (GPR) measurements. This sequential hydrogeophysical inversion approach was applied to data of three wheat crop seasons of a rhizotron facility. The rhizotron facility contains of three plots with different surface water treatments: sheltered, natural and irrigated. This setup provides for each plot a unique data set, which includes root observations, GPR, and soil water potential (SWP) data, at six different depths between 0.1 m - 1.2 m. Above-ground measurements include leaf area index, and consistent atmospheric conditions. The SWC was derived from the horizontal crosshole GPR measurements by analyzing the first arrival times of the electromagnetic waves along a pair of tubes. The results indicate, the SWC varies horizontally and vertically depending on weather conditions, soil properties, and root growth.

In the next step, we established a one-dimensional hydrological model using the software HYDRUS‑1D. It considers vegetation data, soil information, and atmospheric conditions. For the starting model, we used the soil hydraulic parameters derived by fitting the soil water retention curve to the SWP and the GPR SWC data, for all plots. For the sequential hydrogeophysical inversion, we combined the hydraulic model with the geophysical data set, for the natural plot. First, we used synthetic GPR data, to proof the concept of the inversion approach. Secondly, we applied the approach to field data.

This study illustrates the feasibility of using time-lapse horizontal borehole GPR data to determine soil hydraulic parameters. These developments are essential to estimate the potential to use geophysical methods that enable non-destructive SWC-observation at the field-plot scale with a high spatial and temporal resolution.