ASSESSMENT OF ENVIRONMENTAL CHANGE AND LAND DEGRADATION USING TIME SERIES OF REMOTE SENSING IMAGES

Abstract

Remotely sensed images having different spectral, spatial, radiometric resolutions provide up-to-date and repetitive information about the Earth's surfaces. This valuable information is essential to determine and monitor environmental changes for decision makers dealing with environmental management. Uncontrolled and rapid urbanization has been one of the major problems faced in the development of the countries, and has caused land degradation resulting in serious and unexpected environmental problems. The aim of the study is to determine land use and land cover changes in a highly industrialized and urbanized region, Kocaeli province of Turkey, using multitemporal satellite images for the period of 1984-2009. Five dates of Landsat images from 1984, 1987, 1999, 2002 and 2009 were selected to conduct multitemporal change detection. Based on the supervised classification method of support vector machines algorithm, images were classified and then post classification comparison technique was used to analyze the changes occurred during the 25-year period. It was found that rapid urbanization occurred in the study area with an increase of 184% in urban lands, and forests in the region are threatened by the high-level of urbanization especially after the year of 2002.