Assesing different remote sensing techniques to detect land use/cover changes in coastal zone of Çukurova deltas

This study assessed different land cover change detection algorithms in Cukurova Deltas, south Mediterranean coastal zone of Turkey using remotely sensed data and Geographic Information Systems (GIS). Four algorithms were used including: image differencing, image rationing, image regression and change vector analysis. Change detection error introduced by seasonal differences was minimized using close dates of Landsat TM data recorded in 7 July 1985, July 1993 and 21 July 2005. Initially, images were geometricaly corrected to the utm projection system, then, atmosferic correction and radyometric normalization was applied to the images. Following to this pre-processing applications, change detection techniques were applied. Object based maximum likelihood classification was used with crossclassification technique to determine `from-to? change which made a comparasion among different techniques using accuracy assessment . This results indicated a decision about most suitable technique for the study area. Sand dunes conversion to agricultural land and coastal erosion were most significant changes in the study area. In practice, different algorithms are often assessed to find suitable change detection approach for a specific application.