Extracting and analyzing forest and woodland cover change in Eritrea based on landsat data using supervised classification

Abstract

Remote sensing images are suitable for quantifying and analyzing land-cover dynamics, particularly for forest-cover change. In this study, the methodology used the supervised classification technique to classify and analyze the total forest-cover change in Eritrea. The results indicated that the forest and woodland cover extracted with high overall accuracy and kappa coefficient of approximately 96% and 0.94, respectively. Generally, the forest cover declined from 2966km2 to 1401km2 from the 1970s to 2014, and the woodland forest cover was reduced from 14,879km2 to 13,677km2 in the same period. The annual rate of deforestation was very high, with approximately 0.35% (62km2) of the total forest cover lost each year for the last 44years. The study concluded that deforestation is one of the leading causes of environmental degradation in the country and it might be caused by human factors as well as due to climate change, i.e., by prolonged drought and inadequate and erratic rainfall. Thus, this paper may significantly help decision makers and researchers who are interested in remote sensing for forest management and monitoring, and for controlling and planning development at local, regional, and global [scales].