Prediction of urban land use evolution using temporal remote sensing data analysis and a spatial logistic model

Urban land use systems are complex systems with components, factors and agents from natural, environmental, social and economic systems. In this paper, we developed a remote sensing and GIS-based integrated approach to modeling and predicting spatially-explicit urban land use changes. The model was built using temporal remote sensing data land use analysis coupled with a Markov model and a spatial multinomial logistic regression framework. Experiments were performed in the Shenzhen Special Zone to substantiate the accuracy of the proposed method. We show that integration of a Markov model and a spatial logistic model is an effective method to describe urban land use evolution and meet the needs of land use early warning and annual land supply planning.