Response of Simulated Upflow from Shallow Water Tables to Variations in Model Parameter Values

Abstract

Listen

Translate article

Shallow water tables are a widespread and deleterious feature of irrigated landscapes, but they can supply a substantial proportion of irrigated crops’ water needs. Better information on uptake from shallow water tables is required to devise improved irrigation management strategies which prevents the unnecessary leaching that causes water tables to rise. This study assessed the sensitivity of a numerical Richards’ equation-based model to variations in values of input parameters, to better understand the limitations of such models for assessing irrigation management strategies in the presence of shallow water tables. Three soils (clay loam, light clay and clay) were used in the simulations. Water table depths were set at 1.2 m and two different water application regimes (14 or 28 mm applied on the first day of the week) were simulated. The simulations were run over 24 weeks to represent periods prior to plants being present, during early plant growth development and when plants had achieved full growth and canopy development. Simulated upflow rates ranged from~ 2 mm week-1 prior to plant development to 6–25 mm week-1 after full plant development, depending on soil texture and water application. Highest upflow rates occurred in the lighter textured soils and with the low water application. Sensitivity of capillary upflow to variations in model parameter values was assessed from the weekly upflow occurring under a range of parameter values (± 50% of the median value of the soil). Simulated upflow was most sensitive to saturated water content and air entry potential before and during plant development, and to root length density near the water tables after plants were fully developed. In general however, variations in upflow were smaller than the corresponding variations in the parameter values (ie, a 50% change in the parameter value produced a ≤ 50% change in upflow). There were two exceptions to this result, in the clay soil after full plant development and in the clay loam as plant development was initiated. In these situations variations in upflow were double the variations in root length density and air entry potential parameters, respectively. This study showed the importance of root function in determining water fluxes in the presence of shallow water tables. Models to be used for examining management practices in these situations will require a description of root growth appropriate for the context of the study. While the simple description of root development and distribution, especially with respect to the water table, used in this study provided sensible upflow rates, more detailed descriptions of root growth would be required to simulate a broad range of conditions accurately