Estimation of soil hydraulic parameters for fine-textured soils using HYDRUS-1D coupled with PEST

Developing strategies for agricultural water management requires robust evaluation of soil hydraulic parameters. In this study, HYDRUS-1D coupled with Parameter ESTimation (PEST) software was used to estimate the van-Genuchten-Mualem soil hydraulic parameters for a fine-textured soil in the Carman-Elm Creek area, Manitoba, Canada over four growing seasons (2016 – 2019). 100 initial parameter sets were generated per season for calibration using field-measured volumetric soil water content within the 0 – 10 cm, 10 – 30 cm, 30 – 70 cm, and 70 – 130 cm soil layers. Coupling HYDRUS-1D with PEST revealed the non-uniqueness of the solutions, as multiple parameter sets yielded comparable model performance. Sequential calibrations with randomized initial parameter sets and performance-based filtering helped to identify stable, maximum-likelihood parameter ranges. Model performance based on the Nash-Sutcliff efficiency coefficient (0.65 – 0.92), percent bias (-9.31 % - 1.48 %), ratio of root mean square error to standard deviation (0.29–0.59), and the coefficient of determination (0.68 – 0.96) was satisfactory for all seasons except for the validation for 2018 due to high variability between calibration and validation periods within the 2018 season. The saturated hydraulic conductivity showed substantial year-to-year variation (120, 10, 21, and 120 cm day−1 for 2016, 2017, 2018, and 2019, respectively) within the top 10 cm layer and remained relatively stable (< 20 cm day−1) in deeper layers. These findings underscore the importance of accounting for soil-plant-weather interactions when characterizing soil hydraulic parameters to improve agricultural water management strategies and support adaptation in cold-climate cropping systems.