A SWIR-based vegetation index for change detection in land cover using multi-temporal Landsat satellite dataset

Remotely sensed spectral vegetation indices (VIs) are widely used and have benefited numerous disciplines interested in the analysis of biomass, vegetation, plant stress, plant health, crop production, and change detection. This study’s objective is to propose a vegetation index based on short wave infrared (SWIR) for detecting the changes in geographical features using multi-temporal and multi-spectral satellite imagery. The proposed a vegetation index SWIR based normalized difference vegetation index (SNDVI) can play a significant role in monitoring vegetation plant’s health and growth and also increase the accuracy of classification and change detection. In this study, Landsat satellite imagery of different periods (1996, 2003, 2010 and 2017) has been used to develop the vegetation index. The proposed vegetation index is useful for increasing the accuracy of classification and change detection of geographical features. Using the Normalized Difference Vegetation Index (NDVI) method, the accuracy of classification of orchards, vegetation and rangelands has been obtained at 96.90%, 94.68% and 92.29% in 2017. While using the SNDVI method, higher accuracy of 98.09%, 95.50% and 92.02% has been obtained in the classification of orchards, vegetation and rangelands. The proposed SNDVI method can be significant in monitoring the health and growth of vegetation plants and detecting the changes in orchards biomass and other vegetation plants.