Simulation of spatial-temporal change of urban land use/cover in a short period of time

Abstract

The objective of this study is to simulate and validate temporal and spatial changes of land use/cover in a short time period based on the Markov cellular automata (Markov-CA) model that combines Markov chain analysis and cellular automata models. Aim to this task, this paper first analyzed the land use/cover change (LUCC) of an urban area in 1997 and 2000 and computed transition probabilities using Markov chain. Then Markov-CA model was employed to simulate urban land use/cover in 2002 and a validation was followed by using actual land use map interpreted from the satellite image captured in the same year. Experimental result shows this model is reliable with an acceptable overall accuracy (86.52%) and Kappa coefficient (0.79), which suggests that LUCC in urban within a short time period could be modeled and predicted by Markov-CA model.