Deforestation Analysis of Northern Areas(Pakistan) using Image Processing and Maximum Likelihood Supervised Classification

Abstract

An extensive uncertainty is a big concern of deforestation, degradation and forest decentralization. Recorded cases of deforestation in northern regions of Pakistan have drawn serious involvement in the last two decades. These areas include agriculture land and source of fresh water for more than 20 million residents. Downgrading in the forest is also big damage in the ecosystem which increases the flood risk in any community. The fast development in remote sensing (RS) satellites and RS techniques in the last four decades, provides a stable, successive and efficient way for analysis of land cover and land mapping. This work is carried out the analysis of deforestation from 2000-2016. The satellite images of Landsat 4, 5 and 8 are used and processed by Arc GIS software. Maximum Likelihood (ML) supervised classification based on Bayes theorem is used and this technique classified the vegetation, water, and mountains individually. Normalized Difference Vegetation Index (NDVI) is used to calculate the downfall in forests from the vegetation area. This analysis reflected a compelling downturn in forest cover in a period of study.