An integrated approach of logistic-MCE-CA-Markov to predict the land use structure and their micro-spatial characteristics analysis in Wuhan metropolitan area, Central China.

As human interference with the natural environment accelerates, land use has undergone great changes. However, to realize rational land development in the rural-urban ecotone, the micro-spatial (MS) unit is the best scale for the management and planning of sustainable land use. Taking Wuhan metropolitan area as research area, the integrated logistic-multi-criteria evaluation (MCE)-cellular automata (CA)-Markov model was used to simulate land use pattern for 2025. In addition, the 1 km×1 km, 2 km×2 km, 3 km×3 km, and 4 km×4 km and typical sample belt were built to reveal the spatial microcosmic expression of land use structure. The results showed that the kappa coefficient and figure of merit (FoM) were 88.01% and 26.86%, respectively, indicating the integration model has high prediction accuracy. In 2005-2025, the diversification of land use in the Wuhan metropolitan area will be generally above the medium level, and the types of land combinations will be relatively abundant. As human activities increase, the land use degree will show increases continuously, it will expand outward from Wuhan, and there is a positive correlation between cultivated land-rural residential land and urban land-cultivated land. The spatial distribution of land use structure presents regional scale characteristics, and different regions have micro-spatial scale dependence. The selection of MS scales based on local conditions can be a good way to reflect land use internal structure and provide a better reference for the compilation of regional land use optimization.