农业机械学报2018，Vol.49Issue(1)：151-156,6.DOI:10.6041/j.issn.1000-1298.2018.01.019
基于Sentinel-2遥感影像的玉米冠层叶面积指数反演
Retrieving Leaf Area Index of Corn Canopy Based on Sentinel-2 Remote Sensing Image
摘要
Abstract
Leaf area index is one of the important parameters to describe the canopy structure of corn,which determines the biophysical processes of corn canopy photosynthesis,respiration,transpiration and carbon cycle.Therefore,retrieval of leaf area index is of great significance to corn growth monitoring.The Sentinel-2 remote sensing image and LAI-2000 ground synchronous data were used to retrieve the leaf area index of corn canopy.Normalized difference spectral index (NDSI) and ratio spectral index (RSI) were extracted to build the univariate and multivariate empirical models.The best LAI retrieving models were identified based on the best combinations of coefficient of determination (R2) and root mean square error (RMSE).Finally,spatial distributions of LAI in the study area were mapped through the optimal retrieve model.Results showed that all spectral indices tested were significantly correlated with LAI of corn,and the correlation between spectral indices built with red-edge bands and LAI was higher than that built without red-edge bands.Validation analysis result indicated that although the accuracy of the multivariate empirical model was high,its ability to predict LAI was poor.Linear regression model of NDSI(783,705) most accurately explained retrieval of LAI of corn,with R2 of 0.534 2 and RMSE of 0.288 5.Therefore,linear regression model of NDSI(783,705) was recommended as the most legible model for estimating LAI of corn.The rededge bands confirmed from Sentinel-2 remote sensing image improved the accuracy of retrieving the LAI of corn.Moreover,the results also provided a powerful evidence to develop the Sentinel-2 remote sensing image and red-edge bands application in retrieving the LAI of corn.关键词
Sentinel-2遥感影像/玉米冠层/叶面积指数/红边波段/光谱指数Key words
Sentinel-2 remote sensing image/corn canopy/leaf area index/red-edge bands/spectral indices分类
信息技术与安全科学引用本文复制引用
苏伟,侯宁,李琪,张明政,赵晓凤,蒋坤萍..基于Sentinel-2遥感影像的玉米冠层叶面积指数反演[J].农业机械学报,2018,49(1):151-156,6.基金项目
国家自然科学基金项目(41371327、41671433) （41371327、41671433）
