Search

Two promising techniques for estimating Leaf Area Index (LAI) using remote sensing are Linear Spectral Mixture Analysis (LSMA) and Modification of Spectral Vegetation Indices (MSVI). The Normalized Distance Method (ND), which uses principles employed by the LSMA and MSVI techniques, is introduced in this study. These three methods are applied to a region of montane forest in Kananaskis Country, Alberta, Canada, in order to estimate LAI. In situ measurements of LAI in 10 deciduous and 10 coniferous plots, and a SPOT-4 image taken at the height of the growing season, provided test data that produced relationships for LAI in pure stands of either coniferous or deciduous vegetation using each of the three methods. All methods exhibited varying degrees of performance and demonstrated significant dependence on vegetation type. The ND method produced relationships with coefficients of determination (R 2) of 0.86 and 0.65 for coniferous and deciduous vegetation, respectively; the MSVI method (when using the adjusted Normalized Difference Vegetation Index) produced relationships with R 2 values of 0.79 and 0.59 for coniferous and deciduous vegetation, respectively; and the LSMA technique produced relationships with R 2 values of 0.83 and 0.0 for coniferous and deciduous vegetation, respectively. [ABSTRACT FROM AUTHOR]