MODELLING LAND-USE CHANGE DYNAMICS IN A MEDITERRANEAN COASTAL WETLAND USING CA-MARKOV CHAIN ANALYSIS

Rapid development in agricultural activities has resulted in a significant amount of habitat loss at ecologically sensitive areas. This is especially true in coastal regions of the Mediterranean, such as Cukurova Delta. This study aims at combining satellite remote sensing, GIS, and Cellular Automata (CA) Markov chain modeling to analyze and predict land-use/cover changes in Cukurova Delta, Turkey by 2023. Four remotely sensed images recorded in 1977 (CORONA), 1987 and 1998 (SPOT), and 2007 (ALOS) were used to derive land-use/cover information to calculate transition probability matrix. Spatial variables including distance from lagoon, distance from coastline, soil texture, soil groups and land-use/cover suitability maps were derived in a GIS environment. Maps of suitability were generated for each land-use class using the spatial variables together with multi-criteria analysis. Various combinations of image pairs were used to drive transition probability matrix. 2023 land-use/cover change map was based on the change between 1977 and 2007. The land-use/cover change matrix between 2007 (actual) and predicted 2023 maps were created, and change detection analysis was performed. Most remarkable change was an increase in agricultural areas by 10283 ha between 1977 and 2007, with 1766 ha projected to be converted to agricultural use.