Method for information mining of land-use change and its applicaion

Methods for land-use net change, swap change, total change information assessing were constructed based on land-use transformation matrix with the crisp land-use classification scheme. Analysis of the relative differences between the observed transition and its expected transition percent of assumed random transition can extract the clearly systematic inter-category transition. A multiple resolution and soft land-use classification algorithm provides additional information about space distance over which the swap change occurs. These methods were used in the 1991～2001 land-use change analyses in Haidian district, Beijing. The application shows that these methods enable land-use researchers to find the systematic land-use transition information, which can be useful in land-use change analyses and modeling.