Impacts of measured soil hydraulic conductivity on the space–time simulations of water and nitrogen cycling

Abstract

In agriculture, variations in soil nutrients and water are driven by soil properties, topography and agronomic practice that typically interact and change over space and time. Agroecosystem models need to capture these sources in variation, where this study’s first objective was to assess the potential of using measured saturated soil hydraulic conductivity (ksat) to improve the simulation accuracy of water and soil mineral nitrogen content from the SPACSYS model for a lowland UK grazed field (6.34 ha). As a second objective, SPACSYS was run at the field level and at the within-field level to provide a further comparison of simulation accuracy. For model calibration, ksat was measured at 27 points at 0–10, 10–20 and 20–30 cm soil depths on a 50 × 50 m grid. For model validation, moisture and mineral nitrogen content in the same three soil layers, at 10 adjacent points on a 25 × 25 m grid, were measured monthly from May 2018 to April 2019, together with in situ field level water flux measurement. Measured ksat coupled with the within-field setting allowed a novel spatial investigation of SPACSYS performance. Measured ksat (as opposed to unmeasured, default values) was found to improve water flux simulation, but only slightly so, which was considered in part due to a high positive skew in the measured ksat coupled with no clear spatial structure. Field level and within-field specifications simulated soil moisture with equal accuracy, while simulation accuracy of soil ammonium and nitrate improved via the within-field setting; for water flux simulation, the field level setting should be preferred. Results provide further evidence for when a field level setting should be preferred to a within-field setting and vice-versa.