Abstract: Taking Poyang Lake region as an example, we applied the Logistic regression to calculate the driving effects between various factors and land-use types. A Cellular Automata model combining the Markov method was built to simulate the land-use change in the Poyang Lake region. We completed the precision tests of the proposed model and predicted the developing trends of the land-use change in the region. The results show that: 1) The variation of land use types in Poyang Lake region had an obvious ascending trend from 2000 to 2010.  Grassland was the most active land-use type. The area of cultivated land had sharply declined, and the expansion of the residential land kept stable. 2) The driving directions and the explanatory abilities of different factors had significant differences on influencing land-use change. The overall explanatory effect was good and all the ROC values were higher than 0.81. 3) The overall Kappa index for the proposed CA-Markov model was higher than 0.85 and the overall precision index was over 75%. It indicated that the model had a high-precision simulation effect and had a good performance in simulating the changes of woodland, cultivated land and water. 4) The variations of the land-use types in Poyang Lake region increased from 2010 to 2025, whereas their active levels had a stable trend. 5) The lost area of cultivated land and woodland was mainly caused by the gain area of the residential land from 2010 to 2025, the land-use change pattern performed a significant one-way transmission trend from other land-use types to the residential land. The regional urbanization accelerated obviously and urban expansion mainly distributed in the Nanchang-Jiujiang metropolitan area. 6) The lost area of grassland and the unused land mainly distributed in the boundary areas of the counties from 2010 to 2025. The regional urbanization had a positive effect on grassland protection, however, it is easy to cause the inefficient use of cultivated land and the woodland.
Key words：land-use change; Logistic regression; CA-Markov; Poyang Lake region