Improving crop crop yield estimation by assimilating LAI and inputting satellite-based surface incoming solar radiation into SWAP model
Improving crop crop yield estimation by assimilating LAI and inputting satellite-based surface incoming solar radiation into SWAP model
复制标题
DOI：
10.1016/j.agrformet.2017.12.250
复制
发表时间：
2018-03-15
影响因子：
6.2
通讯作者：
Vazifedoust, Majid
中科院分区：
文献类型：
--
作者：
Mokhtari, Ali;Noory, Hamideh;Vazifedoust, Majid
在区域范围内准确预测作物产量将提高全球粮食安全，特别是小麦和大麦等战略作物。土壤水-大气植物模型（SWAP）是一个基于作物生长详细模块的农业水文模型，它可以利用卫星观测数据作为输入数据正确地估计作物产量。在这项研究中，为了减少作物产量估计误差在小麦和大麦，基于MODIS的叶面积指数（LAI）同化使用顺序更新算法到SWAP，和GLDAS/诺亚派生的表面入射太阳辐射（SISR）被用作替代测量SISR。在9个不同的情况下，遥感叶面积指数的同化和使用SISR作为输入进行了研究。结果表明，土壤调节植被指数（SAVI）是估算叶面积指数的最佳方法，其决定系数（R-2）为0.72，均方根误差（RMSE）为0.87m2。噪声等效变化也表明SAVI沿着整个LAI变异范围的适当灵敏度。GLDAS/Noah推导的SISR与实测SISR表现出良好的一致性，因此LAI和SISR联合用于模型中。模拟结果表明，在最大叶面积指数（LAI）达到前10天和最大叶面积指数（LAI）达到后10天的同化过程中，地上干生物量和籽粒产量的绝对误差百分比（PAE）最低，分别为1.59%和6.06%。作物产量估计分别比不同化LAI的情况提高了26.25%和14.4%。总体而言，LAI同化到SWAP与本研究中最有效的情况下，将导致在小麦和大麦作物产量预测准确。
Precise crop yield forecast at regional scales would increase global food security, especially in strategic crops such as wheat and barley. Soil Water Atmosphere Plant (SWAP) is an agro-hydrological model based on a crop growth detailed module that could properly estimate crop yield using satellite observations as input data. In this study, in order to reduce crop yield estimation errors in wheat and barley, MODIS-based leaf area index (LAI) was assimilated using a sequential update algorithm into SWAP, and GLDAS/Noah-derived surface incoming solar radiation (SISR) was used as an alternative to measured SISR. The assimilation of remotely sensed LAI and using SISR as input was examined in nine different cases. Results showed that soil adjusted vegetation index (SAVI) was the best VI for LAI estimation with coefficient of determination (R-2) of 0.72 and root mean square error (RMSE) of 0.87 m(2) m(-2). Also noise equivalent variations indicated an appropriate sensitivity of SAVI along the entire range of LAI variability. GLDAS/Noah-derived SISR showed good agreement with measured SISR; therefore LAI and SISR were jointly used in the model. Simulation results showed that the lowest percent absolute error (PAE) for aboveground dry biomass and grain yield was obtained in case 7 (the assimilation of the peak LAI in addition to ten days after and before the peak LAI is reached) with 1.59% and case 5 (the daily assimilation of LAI until twenty days after the peak LAI is reached) with 6.06%, respectively. Crop yield estimates were improved by 26.25 and 14.4% compared with no LAI assimilation case. Overall, LAI assimilation into SWAP associated with the most efficient cases in this study would result in an accurate crop yield forecast in wheat and barley.