分布式水文模型在徒骇马颊河流域灌溉管理中的应用Ⅰ.参数率定和模拟验证

Application of Distributed Hydrological Model in Irrigation Management of Tuhai-Majia River Basin Ⅰ.Parameter Calibration and Validation

【目的】建立徒骇马颊河流域的分布式水文模型,为进行该地区灌溉制度优化提供支持。【方法】首先运用洗牌复形演化(SCE-UA)算法在禹城试验站和洪门试验站对分布式水文模型(SWAT)中的作物参数进行率定,接着以遥感监测的蒸散数据为目标,应用拉丁超立方-单次单因素(LH-OAT)方法对蒸散发(ET)相关的参数进行了敏感性分析,用序贯不确定性(SUFI-2)算法优化出一套参数范围并进行不确定性分析。在此基础上,在考虑模型不确定性的情况下对冬小麦-夏玉米轮作体系产量的长期模拟结果进行验证。【结果】作物参数率定结果表明,禹城试验站上冬小麦和夏玉米产量的相对误差分别为7.02%和16.60%,洪门试验站上冬小麦和夏玉米产量的相对误差分别为0.09%和0.10%。ET参数率定和统计年鉴产量验证结果:除了NS系数,P_factor、R_factor以及R2的模拟精度较高。【结论】依据徒骇马颊河流域的特点进行SWAT模型相关参数的率定,得到较好的模拟结果,为进一步深入研究该地区水分生产函数和优化灌溉制度打下了基础。

[ Objective] The objective of this study is to establish a hydrological model of Tuhai-Majia river basin which will be a support for irrigation schedule optimization in this region. [Method] Firstly, using ＂Shuffled Complex Evolution＂ （SCE-UA） to calibrate crop parameter of Soil and Water Assessment Tool （SWAT） in Yucheng station and Hongmen station. Secondly, using actual evapotranspiration which calculated by remote sensing to analyze the sensitivity of parameters. Thirdly, using ＂Sequential Uncertainty Fitting Algorithm＂ （SUFI-2） to optimize the parameters and give its uncertainty analysis. Based on the calibrated parameters, considering the uncertainty of the model, the yield of long term simulation of winter wheat and summer maize rotation system was validated. [ Result ] The results of calibration of crop parameter showed that the relative error of yield of winter wheat and summer maize in Yucheng station were 7.02% and 16.60%, and in Hongmen station were 0.09% and 0.10%. The results of calibration of ET parameters and validation of yield showd that the accuracy of P factor, R factor and R2 were reasonable. [ Conclusion] The good results of calibration and validation laid a firm foundation for the further research on water product function and irrigation regime optimization.