Measurement and comparison of Leaf Area Index estimators derived from satellite remote sensing techniques

Abstract

Leaf Area Index (LAI) is an important biophysical characteristic of vegetation that is directly related to rates of atmospheric gas exchange, biomass partitioning, and productivity. Mapping and monitoring LAI over scales from landscapes to regions is essential for understanding medium-scale biophysical properties and how these properties affect biogeochemical cycling, biomass accumulation, and primary productivity. This study developed and verified several models to estimate LAI using in situ field measurements, Landsat Thematic Mapper imagery, vegetation indices, simple and multiple regression, and artificial neural networks (ANNs). It was shown that while multiple band regression and regression with individual vegetation indices can estimate LAI, the most accurate way to estimate regional scale LAI is to train an ANN using in situ LAI data and remote sensing brightness values.