论文部分内容阅读
叶面积指数(LAI)是衡量植被生态状况和估算作物产量的一个重要指标。LAI的反演是定量遥感研究的重要内容。传统的经验统计反演方法基于单一观测角度的遥感数据进行,忽略了地物反射率的方向性。若在反演中加入多观测角度的信息,则有可能提升LAI反演的精度。以2008年甘肃省张掖市玉米实验区为研究区,利用欧空局的CHRIS/PROBA多角度高光谱数据对比分析了传统植被指数NDVI、RVI、EVI的变化规律及其反演玉米叶面积指数LAI的精度,并根据NDVI随观测角度的变化规律,构造出新型多角度归一化植被指数MNDVI,分别对实测叶面积指数进行线性回归并利用实测数据对估算LAI进行精度验证,结果表明:新型MNDVI指数相比于传统NDVI、RVI、EVI对LAI的反演精度有了显著提升,估算模型决定系数R2达到0.716,精度验证均方根误差为0.127,平均减小了33.3%。
Leaf area index (LAI) is an important indicator of ecological status of vegetation and the estimation of crop yield. The inversion of LAI is an important part of quantitative remote sensing research. The traditional empirical statistical inversion method is based on remote sensing data from a single observation point, ignoring the directionality of the reflectivity of the object. If multiple observations are added to the inversion, it is possible to improve the accuracy of the LAI inversion. Based on the CHRIS / PROBA multi-angle hyperspectral data of ESA, the variation regularity of NDVI, RVI and EVI of traditional vegetation indices and their LAI According to the variation rule of NDVI with observation angle, a new multi-angle normalized vegetation index MNDVI was constructed. The measured leaf area index was linear regression and the LAI was validated by the measured data. The results showed that the new MNDVI Compared with the traditional NDVI, the inversion accuracy of LAI is significantly improved by RVI and EVI. The estimated coefficient R2 of the estimated model reaches 0.716 and the root mean square error of precision verification is 0.127, with an average reduction of 33.3%.