Modelling vegetation health and stress using hypersepctral remote sensing data

Abstract

Vegetation health mapping is extremely helpful for distinguishing nuisance and scourge conditions and it is additionally valuable in surveying zones of lumber gather. Decline in forest health occurs due to many reasons, including land-use intensification, forest disturbances and bad management practices at various scale that can be assessed and monitored using geoinformatics. Among various remote sensing (RS) techniques, the use of hyperspectral band information suits more for vegetation stress and health analysis and possesses better accuracy. Besides, RS gives lot of opportunity to monitor and manage natural resources at multi-temporal, multi-spectral and multi-spatial level. In this study, we evaluated the applications of hyperspectral RS for forest health assessment and mapping in terms of vegetation health analysis, stress and fire fuel vegetation analysis and linked these to forest management. We analysed the vegetation health of the Lonar forest (Maharashtra) using hyperspectral imagery using ENVI software for the management of forest and its conservation in terms of healthy and stressed vegetation using different indices. The background studies suggest that a forest displaying low pressure conditions comprise sound vegetation, although a forest under high pressure conditions gives the indication of dry and kicking the bucket plant material, exceptionally thick and inadequate overhang and wasteful light use. We examined different vegetation models to compare and identify the index which resulted showing the best accuracy. We executed six Vegetation Index models (VIM) i.e. Enhanced Vegetation Index (EVI), Sum Green Index (SGI), Moisture Stress Index (MSI), Structure Insensitive Pigment Index (SIPI), Atmospherically Resistant Vegetation Index (ARVI) and Normalized Difference Vegetation Index (NDVI) for evaluation of the timberland strength of the woods locale. The study suggest that each vegetation model index has its specific expression for healthy and stressed vegetation and its own suitability for specific uses and also some limiting factors. We argue on the importance to select an appropriate vegetation model to evaluate vegetation health and stress that must be based on environmental conditions, forest composition, leaf type and other biophysical factors.