摘要
【目的】基于河北省曲周试验站的试验数据,对DNDC模型的参数进行详尽的灵敏度分析和标定。【方法】采用独立参数扰动法进行参数的灵敏度分析,运用试错法对参数进行标定。【结果】在曲周试验站,标定模拟的土壤剖面体积含水量动态的精度较高,平均归一化均方根误差(NRMSE)为22.41%。对于土壤氧化亚氮释放量的模拟,则呈现出明显的季节变化趋势并有效地捕捉到其峰值所在。模型标定模拟的冬小麦叶面积指数、地上部分生物量和产量的NRMSE分别为27.04%、30.46%和28.10%;标定模拟的夏玉米叶面积指数、地上部分生物量和产量的NRMSE分别为28.49%、25.15%和26.92%。【结论】以上标定模拟的精度就田间实际情况而言,总体是可以接受的。经过详细参数标定后的DNDC模型将为深入探讨不同农田管理措施下,实现作物增产和农业减排双赢局面提供科学有效的途径。
[ Objective ] Based on a large number of field measured data in Quzhou Agricultural Experiment Station, detailed sensitivity analysis and calibration of DNDC model parameters were conducted. [ Method ] The independent parameter perturbation method was adopted to carry out parameter sensitivity analysis and using trial and error method was used to calibrate parameters. [ Result ] In Quzhou Agricultural Experiment Station, the accuracy of the simulated soil volumetric water content was higher after calibration, the average normalized root mean square error （NRMSE） value was 22.41%. It showed a clear seasonal trend and captured its peak for the simulation of soil nitrous oxide emission. The NRMSE of leaf area index, aboveground biomass and yield of winter wheat were, respectively, 22.04%, 30.46% and 28.10%. The NRMSE of leaf area index, aboveground biomass and yield of summer maize were, respectively, 28.49%, 25.15% and 26.92%. [ Conclusion ] In a word, the above mentioned simulation accuracy is acceptable. After the detailed parameter calibration of DNDC model, it will provide a scientific and effective way to realize both grain production and agricultural emission reduction under different cropland management measures.
出处
《中国农业科学》
CAS
CSCD
北大核心
2013年第13期2695-2708,共14页
Scientia Agricultura Sinica
基金
国家自然科学基金重点项目(51039007)
水利部公益性行业项目(200901083)
关键词
曲周试验站
DNDC模型
灵敏度分析
参数率定
Quzhou Experiment Station
DNDC model
sensitivity analysis
parameter calibration
