Land use change information mining in highly urbanized area based on transfer matrix: A case study of Suzhou, Jiangsu Province

This paper makes a deep analysis of land use transfer matrix and transfer probability matrix.It puts forward that consecutive time periods of land use transfer matrix can be used to calculate the total time periods of the transfer matrix,also improves the calculation method of land use change and the computational model of land use dynamic degree,further proposes the concept of land use transfer disorder degree and constructs the computational model.Taking Suzhou as an example,this study conducts an empirical analysis of the above models.The results show that: from 1999 to 2008,the total change of cultivated land is the largest,followed by the other agricultural land,independent industrial and mining land.In the composition of the total change,the major change of cultivated land and independent industrial and mining land is the net change,while the major change of the other agricultural land is the swap change.In this decade,the top three of land use dynamic degree are independent industrial and mining land,agricultural land and town land,total dynamic degree is the highest in 2004 and 2002,and less from 1999 to 2001.The top two of land use transfer disorder degree are the other agricultural land and transportation land and the values of other land use types are low.This can reflect that 10 years' land use in Suzhou is in an orderly evolution process.The results show that the improvements and construction of model based on the transfer matrix can explore the information of land use evolution,and also be more conducive to the research of land use changes.