基于WOFOST模型与遥感数据同化的县级尺度玉米估产研究  

Maize Yield Estimation at County Level Based on World Food Studies Model and Remote Sensing Data Assimilation

作　　者：钱凤魁[1] 王化军 王祥国 于远俊 辛家佶 顾汉龙 QIAN Fengkui;WANG Huajun;WANG Xiangguo;YU Yuanjun;XIN Jiaji;GU Hanlong(College of Land and Environment/Key Laboratory of Three-dimensional Protection and Monitoring of Cultivated Land/National Engineering Research Center for Efficient Utilization of Soil and Fertilizer Resources,Shenyang Agricultural University,Shenyang 110161,China)

机构地区：[1]沈阳农业大学土地与环境学院/耕地立体保护与监测重点实验室/土肥资源高效利用国家工程研究中心,沈阳110161

出　　处：《沈阳农业大学学报》2024年第2期138-152,共15页Journal of Shenyang Agricultural University

基　　金：国家自然科学基金联合基金项目(U23A2053)。

摘　　要：区域尺度的作物生长动态监测和产量预测对于保障粮食安全和农业政策的制定具有重要参考依据。遥感数据同化应用极大提高了作物估产的时效性和精度。为及时、准确地实现县级尺度粮食产量的估测,以及提升产量估测的精度,以辽宁省铁岭县为研究区,采用WOFOST(world food studies)模型与遥感同化相结合的方法对铁岭县玉米进行估产研究。通过采用扩展傅里叶幅度敏感性检验算法(extened Fourier amplitude sensitivity test,EFAST)敏感性分析方法实现玉米估产敏感性参数的分析,以及本地化;通过采用参数自动率定程序PEST(parameter estimation)实现参数的优化,验证结果为采样点产量的平均误差为852.39 kg·hm^(-2),模型模拟的精度达到92.82%。为进一步提高和优化模型估产精度,将遥感反演得到的叶面积指数采用集合卡尔曼滤波算法与模型模拟的叶面积指数进行数据同化,平均误差从同化前的852.39 kg·hm^(-2)降低为435.01 kg·hm^(-2),估产精度从92.82%提高到96.33%,有效提高了WOFOST模型估产的精度。结果表明:水分对玉米的生长发育限制并不大,其产量形成主要受光温影响,对温度、光能利用效率和最大同化速率有关的参数具有较高的敏感性;优化后的模型能够较好模拟铁岭县玉米生长发育情况,产量验证表明优化后的模型模拟的效果较好,但仍存在一定的误差;比值植被指数与叶面积指数的相关性最高,反演模型精度较好,反演结果表明叶面积指数在抽雄吐丝期差距较大,而在成熟期的差距不大;经过作物模型与遥感数据同化之后,估产的精度得到明显提高,说明遥感与作物模型同化是一种有效地提高作物估产和产量预测精度的方法。Dynamic monitoring of crop growth and yield forecasting at regional scale provide important references for ensuring food security and formulating agricultural policies.The application of remote sensing data assimilation significantly enhances the timeliness and accuracy of crop yield estimation.In order to timely and accurately estimate the grain yield at the county level and improve the accuracy of the yield estimate,Tieling County in Liaoning Province was selected as the research area.The world food studies(WOFOST)model combined with remote sensing data assimilation was employed to estimate maize yield in Tieling County.In this study,the extened Fourier amplitude sensitivity test(EFAST)sensitivity analysis method was utilized to analyze and locate the sensitivity parameters of maize yield estimation.The parameter estimation(PEST)parameter optimization program was adopted to optimize the parameters.The verification results demonstrated that the average error of yield at sampling points was 852.39 kg·hm^(-2),and the accuracy of model simulation reached 92.82%.To further improve and optimize the yield estimation accuracy of the model,the leaf area index obtained from remote sensing inversion was assimilated with the leaf area index simulated by the model using the ensemble Kalman filter algorithm.The average error decreased from 852.39 kg·hm^(-2) before assimilation to 435.01 kg·hm^(-2) after assimilation,and the yield estimation accuracy increased from 92.82%to 96.33%.The yield estimation accuracy of the WOFOST model was effectively improved.The results showed that the growth and development of maize were not limited by water,and the yield was mainly affected by light and temperature,and the parameters related to temperature,light use efficiency and maximum assimilation rate were highly sensitive.The optimized model can simulate the growth and development of maize in Tieling county.The yield verification shows that the optimized model has a good simulation effect,but there are still some errors.The correlation

关 键 词：world food studies模型 遥感 数据同化 玉米估产 铁岭县 

分 类 号：S513.01[农业科学—作物学]