Application of a Quantitative Approach through Mathematical Inversion of a Vegetation Canopy Reflectance Model for Crop Monitoring Using Satellite Remote Sensing

Abstract

A quantitative approach has been made to retrieve biophysical information of an agricultural crop using multidirectional remote sensing data. The approach consists of retrieving the parameters of the surface bidirectional reflectance model through model inversion against directional data generated by a directional radiative transfer model. Model inversion has been performed applying a non-linear optimization scheme against the simulated directional vegetation reflectance data. The technique provides a reasonably good estimation of the canopy biophysical parameters. Remote sensing data are usually contaminated by different atmospheric effects. As a remedial, standard data correction procedure is available that requires a precise estimation of intervening atmospheric parameters. In this connection, estimation accuracies of biophysical parameters of vegetation have been analyzed with simulated satellite data corrected with improper values of atmospheric aerosol contents. The study reveals that estimation can be significantly influenced by the use of improper values of aerosol content during data correction. Estimation accuracy is higher in the near-infrared as compared to that in the visible for both high and low density canopies. Moreover, accuracy is higher for a low density canopy in comparison to that for a dense vegetation canopy.