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

The amount and spatial and temporal dynamics of vegtional LAI–SVI approaches to circumvent these limitaetation are important information in environmental studtions. 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 indeal 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 estimatand associated pixel values. In the second step, the training 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 fivegetation 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 reSVI 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 LAIdistribution. This approach was applied to Landsat TM SVI equation (with a set of coefficients) that can be apimagery acquired in the semiarid southeast Arizona and plied to remote-sensing images of different surface types. AVHRR imagery over the Hapex-Sahel experimental sites The second approach involves using bidirectional reflecnear Niamy, Niger. The results were compared with limtance distribution function (BRDF) models. It inverts a ited ground-based LAI measurements and suggested that BRDF model with radiometric measurements to estimate the proposed approach produced reasonable estimates of LAI using an optimization procedure. Although this apleaf area index over large areas in semiarid regions. This proach has a theoretical basis and is potentially applicastudy was not intended to show accuracy improvement ble to varying surface types, its primary limitation is the of LAI estimation from remotely sensed data. Rather, it lengthy computation time and difficulty of obtaining the provides an alternative that is simple and requires little required input parameters by the model. In this study, we knowledge of study target and few ground measurepresent a strategy that combines BRDF models and convenments. Elsevier Science Inc., 2000