>> 自然科学版期刊 >> 2018年03期 >> 正文
基于EFAST方法的WOFOST作物模型参数敏感性分析
供稿: 陈艳玲;顾晓鹤;宫阿都 时间: 2018-06-22 次数:

作者:陈艳玲顾晓鹤;宫阿都

作者单位:‍北京师范大学环境遥感与数字城市北京市重点实验室北京师范大学环境演变与自然灾害教育部重点实验室国家农业信息化工程技术研究中心北京师范大学地理科学学部

摘要:作物生长模型广泛应用于作物长势监测和产量预测。为了有效识别作物模型关键参数,减少模型模拟的不适用性,选取河北省藁城市2009—2010年冬小麦作为研究对象,应用扩展傅立叶振幅灵敏度检验法(EFAST),对WOFOST模型26个作物参数进行敏感性分析。结果表明,生育期为0.5和1.0时的比叶面积(SLATB1和SLATB2),出苗到开花期的积温(TSUM1),35℃时叶面积的生长周期(SPAN),20℃下单叶有效光能利用率(EFFTB3)和最大CO2同化率在30℃的校正因子(TMPF4)等6个参数的敏感性指数均大于0.1,说明对产量形成的贡献较大。研究证明,基于扩展傅立叶幅度检验法(EFAST)的敏感性分析对模型修正具有指导意义,可为模型参数"本地化"提供重要依据。

基金:国家自然科学基金资助项目(41671412);国家重点研发计划课题(2017YFB0504102);中央高校基本科研业务费专项资金资助项目;

关键词:敏感性分析;扩展傅立叶幅度检验法(EFAST);WOFOST模型;作物参数;

Abstract:Crop growth simulation models are widely applied in crop growth monitoring and yield forecasting. In order to identify the key parameters of crop model and reduce the non-applicability of the model simulation, the winter wheat of Gaocheng was selected as a case. The extend fourier amplitude sensitivity test was employed to analyze the sensitivity of WOFOST model parameters. The result shows that six parameters have high sensitivity with values were greater than 0. 1. They are the specific leaf area at 0. 5 and 1. 0 growth stages, the accumulated temperature from seedling to flowering stage, the growth cycle of leaf area at 35 ℃, the correction factor of effective light energy utilization at 20 ℃ and the maximum CO2 assimilation rate at 30 ℃. It also indicated that the contribution of the parameters to the yield formation was large. The research demonstrated that EFAST method has significant value in guiding model modification and can also provide an important basis for the localization of model parameters.

DOI:10.16186/j.cnki.1673-9787.2018.03.10

分类号:S126;S512.11

最近更新