Application of Remote Sensing and Geographic Information System for Forest Cover Change Detection

Abstract

Forest cover change detection based on remotely sensed data has been established as indispensible tool for providing suitable and wide-ranging information to decision support system for forest management and sustainable development. In this research an attempt is made to study the change of forest cover in Abugeili forest over 24 years period (1987 -2011), using multi-temporal RS data and GIS based techniques. For this study, LANDSAT TM and ETM+ imagery of 1987, 1999, and 2011 were used and supported by field work through which data were collected and used for image registration, classification process and discussion of the results. ERDAS Imagine, ArcGIS and Microsoft excel software have been used for image processing and data analysis. After performing unsupervised classification on these images, a total of four forest covers classes were identified and mapped. These were water body (WB), sparse forest (SF), high density forest (HDF), and bare soil (BS). The classified images have been used to produce the change detection maps and indices of change dynamics. An NDVI and vegetation change matrices were used for accuracy assessment and verification. The results of change detection analysis revealed that, the forest has remarkable change, from 1987-2011 the HDF was decreased by 34.2%, and 47.16% respectively. A significant forest covers reduction by 84.8% of HDF in year 2011 when compared with the SF in that period. The study and field investigation it has been found that, the problem of forest cover changes are directly linked with anthropogenic activities such as human pressure, new construction, as well as demand of forest products for different purposes. In order to old back the problems of forest cover change and identification of areas under risk of invasion corrective measures were suggested.