Application of a root water uptake model and numerical simulation to walnut trees in arid areas of northwest China

Abstract

Water shortages restrict the sustainable development of agriculture in arid areas of northwest China. Quantitative studies of root water uptake (RWU) and soil water dynamics can guide the optimal regulation of irrigation. We studied walnut trees under drip irrigation in Hongqopo Farm Experimental Base in Aksu Prefecture, Xinjiang. Our aim was to parameterize the root distribution of walnut trees, establish a flexible and practical macroscopic two-dimensional RWU model, and numerically simulate soil moisture. Multiple regression analysis was applied to establish the root length density (RLD) distribution function and RWU model. The finite difference method was applied to the RWU rate inversion solution, and the alternating direction implicit difference (ADI) method was applied to numerically simulate soil moisture. The RLD distribution function accurately described the location of maximum root density and zero root density using four spatial distribution parameters (R2 = 0.86). The RWU model comprised three parameters: transpiration intensity, RLD distribution, and soil moisture distribution. There was a close agreement between simulated and measured RWU rates. The RWU was concentrated at depths of 10–60 cm. The mean absolute error of soil moisture numerical simulation for 16 days was 0.027 cm3·cm-3, which indicated that the proposed RLD distribution function, RWU model, and numerical simulation method were reliable. These approaches have the advantages of simple parameter measurement and flexible consideration of the spatial variability of the RWU rate and are applicable to other fruit trees. They provide a theoretical basis for water-saving irrigation and scientific management of orchards in arid areas.