Simulating urban land-use changes by incorporating logistic regression and CLUE-S model: a case study of Wuhan city

Land use and land cover (LUCC) changes are regarded as the most prominent influence of human activity on the environment. In this paper, a simulation algorithm of land use was developed by combining the logistic regression, Markov model, and CLUE-S model. Based on the observed land use maps and environment dataset, the integrated model was employed to predict the spatial distribution maps of land use from 2015 to 2035 in Wuhan city. The results show that: (1) the simulated maps have good consistency with the observed maps. The accuracy of simulation results exhibit decreased from 2000 to 2015, but the integrated model still achieved acceptable accuracy in 2015, with OA of 78.73% and Kappa of 0.68. (2) During 2015–2040, the built-up area is expanding obviously, most of the expansion area distribute in the center of Wuhan city. The artificial water and cropland tend to decrease and the area of natural wetland and green land remain roughly unchangedfrom 2015 to 2035.