Leaf Area Index Estimates Using Remotely Sensed Data and BRDF Models in a

The amount and spatial and temporal dynamics of veg- tional LAI–SVI approaches to circumvent these limita-etation are important information in environmental stud- tions. The proposed strategy was implemented in three ies and agricultural practices. There has been a great sequential steps. In the first step, a BRDF model was in-deal of interest in estimating vegetation parameters and verted with a limited number of data points or pixels to their spatial and temporal extent using remotely sensed produce a training data set consisting of leaf area index imagery. There are primarily two approaches to estimat- and associated pixel values. In the second step, the train-ing vegetation parameters such as leaf area index (LAI). ing data set passed through a quality control procedure The first one is associated with computation of spectral to remove outliers from the inversion procedure. In the fi-vegetation indices (SVI) from radiometric measurements. nal step, the training data set was used either to fit an LAI– This approach uses an empirical or modeled LAI–SVI re- SVI equation or to train a neural fuzzy system. The best lation between remotely sensed variables such as SVI and fit equation or the trained fuzzy system was then applied biophysical variables such as LAI. The major limitation to large-scale remote-sensing imagery to map spatial LAI of this empirical approach is that there is no single LAI- distribution. This approach was applied to Landsat TM SVI equation (with a set of coefficients) that can be ap- imagery acquired in the semiarid southeast Arizona and plied to remote-sensing images of different surface types. AVHRR imagery over the Hapex-Sahel experimental site