Spatio-temporal modelling of land use changes by means of CA–Markov model

Abstract

Cities and their problems in our world today are very important because they can affect the land uses and landscapes around them and make them change. Researchers have been tried to solve these problems using several methods. One way that can be useful in planning is simulation and prediction land use changes according to the parameters that affect changes in land use over the time. In this study, changes in land use of Bojnord city (the capital city of North Khorasan province), were considered and studied. In this regard, the CA–Markov model was used. In this model, suitability maps are the basis of the land use changes transformation. To produce suitability maps in this study, agricultural lands were paid specially attention and suitability of them has been reduced as much as possible in modeling. The land use maps of two different periods of time were used to calibrate the models. To validate the model, the “Validate” method (which is a statistical method to validate models) was used. The calculated kappa coefficients for both models were over 85%. In next step, changes of land use have been simulated until the year of 2050. Examination of the output maps which are obtained by CA–Markov model shows that the most growth in land use is in the built-up areas. In 2050, the built-up areas would grow 5.3% compared to 2009 and in subsequent periods the growth rate would be 3.5% on average.