Investigation of Land Use/Land Cover change in Mersin using geographical object-based image analysis (GEOBIA)

Abstract

In a rapidly changing and developing world, progressive urbanization with a growing population is inevitable. With the migration of people from rural areas to urban areas, both in Turkey and in the world, growth is continuously occurring even in small cities. Remote Sensing technologies provide fast and reliable data and methods for the examination of urban areas, and with these methods, it allows to produce highly accurate Land Use/ Land Cover (LU/LC) maps to examine the temporal changes of urban areas. This study aimed to interpret the changes in Mersin province in the last six years by using LU/LC maps. In this context, Sentinel-2 images of the years 2015 and 2021 with 10 m spatial resolution, including Mersin city center, were obtained. Highly accurate LULC maps of two different years were produced using geographic object-based image analysis (GEOBIA). During the object-based classification, Level 2 of the CORINE project terminology was used for the LU/LC classes. In this process, open-source geographic data (Open Street Map, Wikimapia) were also included in the classification. The LU/LC changes that were occurred in the study area in the last six years were evaluated by comparing the classification results, the areas of the detailed LU/LC classes were examined and the changes were put forward objectively.