Use of multispectral and hyperspectral remotely sensed data for vegetation species discrimination

Abstract

Yolandi Lutuli Pieterse

Rangeland degradation is a serious hindrance to sustainable development in degraded areas. Mapping and monitoring vegetation species is an increasingly important issue across various fields of rangeland management. Remote sensing technology is a tool for mapping and monitoring vegetation species and it provides timely and relatively accurate information concerning degradation in biological rangeland resources. The objective of this review was to provide precise and essential information relating to the application of both multispectral and hyperspectral sensors as well as to their limitations with regard to mapping and monitoring rangeland degradation based on the abundance and distribution of vegetation species and algorithms used to process remotely sensed data when classifying these species. The abundance and distribution of the different vegetation species can be used to indicate the gradient level of rangeland degradation. It can be concluded, that up-to-date, spatial information and appropriate processing techniques are essential requirements for extracting increaser and decreaser spectral information that can be used for sustainable rangeland management.