Urban Growth Prediction Using Cellular Automata Markov: A Case Study Using Sulaimaniya City in the Kurdistan Region of North Iraq

Abstract

Many cities in the Kurdistan Region have witnessed a rapid change in land use during the last two decades. Geographic information systems (GIS) and remote sensing have been broadly utilized to monitor and detect urban growth prediction. In this paper, three Landsat TM 5 and one Landsat 8 of Sulaimaniya city were used to identify and develop an urban growth map for 1991, 1998, 2006 and 2014. A supervised classification approach was applied; in order to predict urban growth, the Markov chain and CA-Markov models were used. The result demonstrates that validation of CA-Markov to forecast 2006 land cover map is ineffective in reasonably predicting land coverage for this time period; however this model had significant validation for the year 2014 and also has a good forecast power for 2024.

Keywords

Land Use Change/Cover, Urban Growth Prediction, Supervised Classification, Markov Chain, CA-Markov, Validation.