基于SHAW模型的北疆地区不同滴灌年限棉田冻融期土壤水热盐动态模拟研究

Dynamic Simulation of Soil Water, Heat and Salt in Freeze-thaw Stage of Cotton Field with Different Mulched Drip Irrigation Years in Northern Xinjiang Based on SHAW Model

摘要:

为探究SHAW(Simultaneous heat and water)模型在北疆地区长期膜下滴灌棉田冻融期土壤水热盐动态模拟的适用性，本研究选用滴灌起始年限为1998年(21 a)的棉田土壤水热盐实测数据对SHAW模型进行率定，以滴灌起始年限为2006年(13 a)、2008年(11 a)、2012年(7 a)和荒地(0 a)的水热盐实测数据进行验证。模型率定结果表明，随土壤深度增加土壤温度的模拟效果越好；土壤水盐的模拟效果先增强后减弱。模拟土壤温度Nash系数(NSE)、均方根误差(RMSE)和R²分别为0.713 ~ 0.993、0.209 ~ 2.498 ℃ 和0.911 ~ 0.994；模拟土壤水分NSE和RMSE分别为0.824 ~ 0.967和0.009% ~ 0.032%；模拟土壤盐分NSE和RMSE分别为0.609 ~ 0.844和0.001 ~ 0.012 g/kg。模型验证结果表明，随滴灌年限增加模拟效果越好，模拟除荒地20 ~ 60 cm土层土壤温度NSE小于0.600，滴灌7、11和13 a地块各层土壤温度NSE均大于0.600，RMSE介于0.143 ~ 3.213 ℃；滴灌0、7、11和13 a地块模拟的各层土壤水分NSE均大于0.670，RMSE为0.009% ~ 0.057%；滴灌0、7、11和13 a地块模拟的除120~140 cm土层土壤盐分NSE小于0.600，其他各层土壤盐分NSE均大于0.616，RMSE为0.000 ~ 0.016 g/kg。总体而言，SHAW模型适用于北疆地区冻融期长期膜下滴灌棉田的一维土壤水热盐模拟。

Abstract:

In order to investigate the applicability of the SHAW (Simultaneous heat and water) model in simulating soil hydrothermal and salinity dynamics during freeze-thaw period in long-term drip-irrigated cotton fields in the Northern Xinjiang region, the measured soil data of cotton fields with the starting year of 1998 (21 a) for drip irrigation were used to calibrate SHAW model in this study. SHAW model was validated with the data from drip irrigation start years of 2006 (13 a), 2008 (11 a), 2012 (7 a), and barren land (0 a). The results showed that, with the increasing soil depth, the simulation accuracy of soil temperature was better, and the simulation accuracy of soil water salinity was enhanced and then weakened. Nash coefficient (NSE), root-mean-square error (RMSE) and R² of simulated soil temperature were 0.713–0.993, 0.209–2.498 ℃ and 0.911–0.994, respectively. NSE and RMSE of simulated soil moisture were 0.824–0.967 and 0.009%–0.032%, respectively, and NSE and RMSE of simulated soil salinity were 0.609–0.844 and 0.001–0.012 g/kg, respectively. The results of model validation showed that the simulation accuracy was better with the increase in the number of years of drip irrigation, and NSE of simulated soil temperature was greater than 0.600 and RMSE ranged from 0.143 to 3.213 ℃ in all layers of the drip-irrigated plots of 7, 11 and 13 a, except that NSE of soil temperature in the soil layer of 20–60 cm in the barren land was less than 0.600. NSE of soil moisture was greater than 0.670 and RMSE ranged from 0.009% to 0.057% in all soil layers of simulated drip irrigated plots for 0, 7, 11 and 13 a. NSE of soil salinity for 0, 7, 11 and 13 a were greater than 0.616 and RMSE ranged from 0.000 to 0.016 g/kg in all soil layers, except for the120–140 cm soil layer with NSE less than 0.600. Overall, SHAW model is an available tool to simulate the soil water, heat, and salt in cotton fields under long-term mulched drip irrigation during the freezing and thawing period in Northern Xinjiang.