نوع مقاله : علمی - پژوهشی
نویسنده
دانشگاه صنعتی اصفهان
چکیده
کلیدواژهها
عنوان مقاله [English]
Markov chain is a model that use for prediction future conditions based on the rates of past change. The method is based on probability that a given piece of land will change from one cover to another. These probabilities are generated from past changes and then applied to predict future change. The main objective of this paper is to evaluate the validation of Markov model in vegetation cover crown percentage change simulation. In this study, TM images sensor for 1989, 1998 and LISS III image sensor for 2006 from Landsat and IRS satellite from Mouthe wild life refuge were used to provide maps of vegetation cover percentage. In order to prediction of vegetation cover in 2006, Markov chain model were used by vegetation map of 1989 and 1998, during 9 years. Validation of all maps was evaluated. Finally the map for 2006 that had provided by image was used to evaluate the map from Markov model. Kappa for this map was 53 %. Post classification was used to investigate the area with high accurate prediction and error in prediction. Results have shown because of high decrease in vegetation cover in non-core zone and increase vegetation cover in core zone this model couldn’t predict by high accuracy. But this model is valuable in small scale, in order to general view of future. Keywords: Marcove chain, satellite images, change detection, prediction accuracy.