Abstract: Using models is an effective way to simulate and predict the spatiotemporal process of urban expansion. In this paper, the CLUE-S, LTM, and SLEUTH models were used, and combined with GIS technology, to simulate and predict the urban expansion process of Shanghai, with the results compared and analyzed. It was shown that the three models had somewhat different mechanisms in the simulation and prediction. CLUE-S and LTM models predicted the urban expansion demand first, and then, through comprehensively analyzing different affecting factors, calculated the probability of urban expansion to achieve the spatial allocation, whereas SLEUTH model did the simulation and prediction directly, based on the analysis of the historical development of urban expansion. The three models simulation results of the urban expansion in Shanghai in 2005 were highly consistent with the remote sensing monitoring results. SLEUTH model had the highest simulation precision, with the Kappa coefficient being 0.85, suggesting that this model was superior to CLUE-S and LTM models. In 2005-2020, the area of urban expansion in Shanghai would be 207.7-320.87 km², and in 2020, the City’s area would be 1121.96-1235.13 km². In the next 15 years, the urban expansion rate of the City would be 13.85-21.39 km²·a^-1. Each of CLUE-S, LTM and SLEUTH models had its advantages and disadvantages, and to integrate the advantages of the three models and develop an open and comprehensive model would be the tendency of simulating and predicting urban expansion.

Key words: forest type, soil dissolved organic matter, soil microbial biomass carbon,  Cunninghamia lanceolata.