论文部分内容阅读
农业系统模型参数优化存在很高的不确定性,是模型应用研究的重点和难点。该研究利用自动优化程序PEST(parameter estimation software)对根系水质模型(root zone water quality model,RZWQM)中土壤参数(土壤水力学参数和根系生长参数)和作物遗传参数进行了优化,结果表明PEST优化模拟结果明显优于传统试错法的校正结果,且具有较高的参数优化效率。模型参数优化不确定性评价表明校正数据和参数初始值的选择、土壤水力学参数估算方法、不同类型参数间的相互作用以及优化目标方程(误差来源计算)都明显影响模型模拟结果。以上过程中土壤水力学参数优化值差异较小,但其土壤水分特征曲线却明显不同。通过以上评价分析提高了RZWQM相关参数优化结果的可靠性及其模拟功能,降低了模型参数优化的不确定性,为PEST优化其他模型参数提供了重要支持。
There are high uncertainties in the optimization of agricultural system model parameters, which is the key and difficult point of the model application research. In this study, soil parameters (soil hydraulic parameters and root growth parameters) and crop genetic parameters in root zone water quality model (RZWQM) were optimized by PEST (parameter estimation software). The results showed that PEST optimization The simulation results are obviously better than those of the traditional trial-and-error method, and have higher parameter optimization efficiency. Uncertainty evaluation of model parameters shows that the selection of initial values of calibration data and parameters, soil hydraulics parameters estimation, interaction between different types of parameters and optimization of objective equation (error source calculation) all have significant impact on model simulation results. In the above process, the difference of soil hydraulic parameters is small, but the soil water characteristic curve is obviously different. Through the above evaluation and analysis, the reliability and simulation function of RZWQM related parameters optimization results are improved, the uncertainty of model parameter optimization is reduced, which provides important support for PEST to optimize other model parameters.