Study On Change And Trend Forecast Of Land Use Based On GIS And PLS-PP Model

Land use as the underlying surface factors of the global environment, its change will bring direct impact on the global environment. With the intensive research of the global change, there is a growing concern about the analysis and forecasting of the land use and the land-cover change. In this paper, we discuss quantificational about the transformation between the Land-Use and Land-Cover dynamic Change process and the land-use tapes on the basis of the remote sensing (RS) and Geographical Information System(GIS). The research core of this paper is to build the Projection Pursuit Based on Partial Least-Square Regression Model (PLS-PP). Yanji is in the east of Jilin Province, as a typical area was selected to be predicted in the aspect of the land-use conditions in future in this paper. And it will provide constructive information to the economic development and the land-use evaluation. The research results are as follows(1)By extracting the land-use information from the TM image of 1992 and 2007, overlay these images through spatial analysis model, we can conclude the land-use situation of Yanji during 1992 to 2007. The land-use of Yanji city was change greatly during 15 years: the agricultural land reduced by 583.62 ha, the residential land increased by 2274.59 ha, the lawn increased by 2184.07 ha, the water area reduced by 143.97 ha, the woodland reduced by 1297.14 ha, the unused land reduced by 2672.58 ha. Range of lawn and unused land vary obviously and residential area and industrial and mining land changed most.(2)Try to analysis the dynamic change of land-use construction, spatial distribution and the pattern of land-use transfer. Lots of unused land converts to the agricultural land and the residential land. The mainly factors that due to the change of land-use of Yanji City are as follows: the population growth, the increasing investment of construction, the huge numbers of project, the rapid development of estate industry and the speeding-up urbanization process.(3)Summarized the research in progress and the various methods and theory of the trend-predict and the land-use construction. Built the PLS-PP model, predict the land-use change of study region, and find that the values that PLS-PP model predicted is more accurate by compared to Least-squares regression. The real example proved the PLS-PP model is feasible and efficient.