Study On The Estimation Of Forest Leaf Area Index Using Multi-angle Hyperspectral CHRIS Data

Forests are important regeneration resources. As the principal part of terrestrial ecosystems, forests are the largest, the most widely distributed, the most complex structure and the most abundant resources of ecosystem; also it is the most comprehensive resource library in nature, biological gene bank and energy storage regulation library. It has an irreplaceable role on improving environment and maintaining ecological balance.Leaf area index (LAI) is the ratio between a total area of standing vegetation and single leaf area. Leaf area index is a significant vegetation structural parameter in the study of plant ecological and one of the most basic parameter of vegetation canopy structure, it also has become an important forest quantitative evaluation standard. In this paper, the multi-angle hyperspectral CHIRS data is used to inverse leaf area index. Through the analysis of hyperspectral and multi-angle sensor CHRIS, estimate leaf area index. Main content and conclusion include the followings:(1) The preprocessing method for CHRIS data can be found through analysing the CHRIS data. By ortho correction, five CHRIS images have been matched.(2) Use normalizing difference vegetation index (NDVI), simple vegetation index (RVI), modified simple ratio index (MSR) to inverse the conifer forest and the mixed coniferous broad leaved forest's LAI. It showed that RVI correlated best with LAI under the same conditions.(3) Analyzing the relevance of LAI with NDVI, RVI, MSR that were in different bands of the CHRIS data, and found that the coniferous forest and the mixed coniferous broad leaved forest had different sensitivity to different band. The maximum correlation of LAI with NDVI, RVI, MSR of Coniferous forest happened in the red band 10 (middle wavelength:709nm), the R2 was 0.7098; meanwhile, that of mixed coniferous broad leaved forest happened in the red band 8 (middle wavelength:697nm), the maximum R2 was 0.6385.(4) The correlation of LAI with NDVI, RVI, MSR from CHRIS image data with different angles has been studied, it revealed that 0°image of NDVI and LAI had the best correlation.0° image of NDVI of coniferous forest had the best correlation with LAI, the maximum R2 can reach 0.7098;-36°image took the second place, but the correlation in +36°image was not good. At the same time, the maximum R2 of 0°image of NDVI with LAI of mixed coniferous broad leaved forest was 0.6385, that of -36°,+36°image took the second place, respectively.