Comparison of principal component inversion with VI-empirical approach for LAI estimation using simulated reflectance data

被引：33

作者：

Chaurasia, S ^{[1
]}

Dadhwal, VK ^{[1
]}

机构：

[1] ISRO, Ctr Space Applicat, RESIPA, Agr Resource Grp,Crop Inventory & Modeling Div, Ahmedabad 380015, Gujarat, India

来源：

INTERNATIONAL JOURNAL OF REMOTE SENSING | 2004年 / 25卷 / 14期

关键词：

D O I：

10.1080/01431160410001685018

中图分类号：

TP7 [遥感技术];

学科分类号：

081102 ; 0816 ; 081602 ; 083002 ; 1404 ;

摘要：

A simulated canopy reflectance dataset for a total of six channels in visible, near-infrared (NIR) and shortwave-infrared (SWIR) region, corresponding to Landsat Thematic Mapper (TM) was generated using the PROSAIL (PROSPECT+SAIL) model and a range of Leaf Area Index (LAI), soil backgrounds, leaf chlorophyll, leaf inclination and viewing geometry inputs. This dataset was used to develop and evaluate approaches for LAI estimation, namely, standard two-band nonlinear empirical vegetation index (VI)-LAI formulation (using Normalized Difference Vegetation Index/simple ratio (NDVI/SR)) and a multi-band principal component inversion (PCI) approach. The analysis indicated that the multi-band PCI approach had a smaller rms error (RMSE=0.380) than the NDVI and SR approaches (RMSE=2.28, 0.88), for an independently generated test dataset.

引用

页码：2881 / 2887

页数：7

共 23 条

[1] Design and analysis of numerical experiments to compare four canopy reflectance models [J].