Landscape pattern change prediction of Jinhu coastal area based on CA-Markov model

Modeling is a valuable way to understand a process, and establishing simulation model is an effective way to study land use/cover changes. This paper formed 1990, 2000 and 2010 landscape pattern maps as the bases of modeling with the help of erdas and arcgis, then used Kappa index and overall landscape level´s index method to do the model validation, and found that the CA-Markov model could be applied to Jinhu coastal scenario simulation. Finally, the author used the CA-Markov model which took the multi-time interval phase transition rule to carry on scenario simulation for this region. Predicted results were as follows: With the region´s economic development, construction land would expand continually, farmland, grassland and waters would reduce in future, and construction land would play an increasingly important role in the landscape type conversion. Dealing with such a dilemma remains a challenge for the local government.