合理定量描述土壤水盐动态及作物生长过程对于干旱灌区制定适宜的农业用水措施具有重要意义.该文以SWAP(soil water atmosphere plant)模型为基础,采用变活动节点法实现了对土壤融化期的水盐运移模拟,并在根系吸水计算中引入了基于 S 形函数的水盐胁迫计算方法,以修正原 SWAP 模型对根系吸水的模拟.进一步嵌入了参数与输入数据较少且可以模拟作物生长过程及实际产量的 EPIC(environmental policy integrated calculator)作物生长模型,构建了改进的农田尺度土壤水盐动态与作物生长耦合模拟模型— SWAP-EPIC.分别采用宁夏惠农灌区春小麦和春玉米田间试验数据,对 SWAP-EPIC 模型田间适用性进行了检验.对比分析各层土壤水分与盐分浓度、作物生长指标(叶面积指数、地上部生物量)的模拟值与实测值,结果表明：春小麦和春玉米试验中土壤水分的平均相对误差MRE 和均方根误差 RMSE 均接近于0且模型 Nash 效率系数 NSE 值趋近于1,水分模块模拟精度较高,盐分浓度模拟存在略微差异但总体上一致性较好,并且作物生长指标匹配良好；同时,模拟的产量和蒸散发均较为接近实际值,春小麦和春玉米产量模拟相对误差分别为4.9%和3.3%.综上,该文改进的 SWAP-EPIC 模型可良好地应用于寒旱区农田尺度土壤水盐运移与作物生长耦合模拟.%The quantitative description of soil water flow, solute transport and crop growth processes at field scale is significant for the decision-making of appropriate water use practices in arid irrigation districts. In this study, a modified agro-hydrological model (SWAP-EPIC) for coupled simulation of soil water flow, solute transport and crop growth was developed based on SWAP (soil water atmosphere plant) model and EPIC (environmental policy integrated calculator) crop growth model. The variable active-node method was adopted into the original SWAP model for reasonably simulating the soil water and solute transport processes during soil thawing period. Additionally, the S-shaped osmotic head-dependent functions for describing water and salt stress were also introduced. Further, the EPIC crop growth model, which could simulate the crop growth process and actual crop yield with moderate data input and parameters, was coupled into the SWAP model. Then the field applicability of SWAP-EPIC model was respectively tested using the field experiment data of spring wheat and spring maize at Huinong experimental site in Ningxia. The simulated and observed soil moisture, salinity concentration, and crop growth indicator (leaf area index and dry above-ground biomass) were compared for spring wheat and spring maize. The results showed that the soil moisture was matched very well, with MRE (mean relative error) and RMSE (root mean square error) close to zero and NSE (NSE and Sutcliffe model efficiency) approached to one. The simulated and observed salinity concentration showed an agreement with some slight discrepancy. The simulated LAI and above-ground biomass both matched well with observed ones. Meanwhile, the simulated crop yield was also close to the observations, with relative errors of 4.9% for spring wheat, and 3.3% for spring maize. The results indicated that the modified model (SWAP-EPIC) could be efficiently used to simulate the soil water and salt dynamics, crop growth, and their relationships at field scale.
展开▼
