Geospatial hotspot detection of forest loss hotspot using NDVI as a response variable

Abstract

Geospatial hotspot detection is a scan statistic process that identifies a region exhibiting characteristics of interest (unusual, anomaly, outbreak, critical resources area). The hotspot detection is a statistical process. It requires a geospatial dataset as an input, which includes two variables for hotspot detection, Size or Population and Response or Cases. In deforestation hotspot detection total area of forest is a size variable whereas deforested area is a response variable. In this paper we figure out deforestation hotspot using Normalized Difference Vegetation Index (NDVI) as a response variable. The use of NDVI as a response variable in hotspot detection is experimented for the first time. The paper presents and elaborates algorithm that describes the use of NDVI as a response variable.We discuss in detail how NDVI can be used to obtain appropriate vegetation information from a remote sensing raster image (Satellite image). We compare hotspot obtained using NDVI as a response variable with the conventional method of determining deforestation (Land Use Land Cover). NDVI brings automation in data generation and hence it would be free from any bias and speed up the overall process of hotspot detection. The proposed technique shall play a significant role in efficient and effectiveness governance.