Abstract: Transfer balance table was obtained through field investigation and summary by the department of land and resources management. It is of high accuracy on time interval and land classification, but now its intensive application is less in the study on land use change and prediction. The purposes of this study are to explore the application methods of transfer balance table and help information extraction and prediction of regional land use change.Land use transfer balance table can be regarded as land use transfer matrix. Based on the relationship between transfer matrix, transfer probability matrix and land use structure vector, we put forward a method which can derive the transfer balance table of years period with consecutive annual transfer balance table. Meanwhile, we propose the idea of analyzing land use order degree based on total land use dynamic degree. We discuss the method which can predict the land use structure of any year with consecutive annual transfer balance table. Because traditional forecasting method based on average transfer probability matrix has some drawbacks, we deduce prediction probability matrix by calculating the root of total transition probability matrix, and predict the land use structure in the future by using MARKOV theory.Using the above methods, in this paper we analyzed and predicted the land use structure of Suzhou City with annual transfer balance table between the year 1999 and 2008. The results showed that: First, as Suzhou City is in a period of rapid economic growth, the biggest feature of land use change is the transfer from cultivated land to various types of construction land, the whereabouts of the numerous reduction of cultivated land is the source of the increase of various types of construction land. Second, the total land use dynamic degree of the 10 year cycle was 15.66%, the sum of the annual total land use dynamic degree was 17.13%, the difference between the two was little, showing that the land use order degree was high in Suzhou City. Third, we predicted the land use structure of any year and the method above was proved to be feasible. The prediction results showed that by 2020, the trend of land use change was still the reduction of agricultural land and cultivated land and the increase of the total amount of construction land, but the rate of transformation will be gradually reduced. While urban land, town land, independent industrial and mining land will increase, rural residential land will be reduced year after year. The sources of construction land will gradually shift from extending to the periphery to improving the internal potential of construction land. Fourth, the prediction result is that the land use structure in Suzhou City will become stable by the year 2040, which is consistent with the prediction of the stage of economic and social development. The study is significant for the intensive applications of land use transfer balance table.

Key words: transfer balance table, application method, change analysis, prediction, Suzhou