Prediction of the Construction Land Structure Based on an Improved DEA Method:A Case Study of Nantong City
基于DEA无效改进的南通市建设用地结构预测

A certain level of economic development needs a certain structure and scale of construction land. Only based on the realistic economic development can the prediction of the structure and scale of construction land be reasonable. Consequently, unreasonably expanding of the scale of construction land would be restrained. The use of construction land would become intensive in the future, which would be beneficial to protecting cultivated land. However, most of published studies regarding the prediction of the demand for land seem to fall short of examining the efficiency of construction land supply to economic development. Because of only depending on historical data of land use, the scale of construction land in the future would be overestimated. In this study, we projected the scale and demand of construction land in Nantong City, Jiangsu Province. Data about construction land were obtained from an investigation into the alternative land use in Nantong City during the period 1996-2008. Economic data were obtained from Nantong Statistical Yearbook from 1996 to 2008. First, we used the method of DEA to analyze the efficiency of the construction land structure to the level of economic development. Second, we improved the data that were considered invalid by an improved formula of DEA. Third, we used the improved array to project the structure and scale of construction land in 2010 and 2020 by the method of GM (1, 1). It can be concluded that 1) from 1996 to 2008, there were 5 years (i.e., 1998, 1999, 2004, 2005, and 2008) that the efficiency of construction land was essentially invalid; 2) the overall scale of construction land will continue to expand until 2020, but the alternative tendency of each type of construction land varies to certain degrees. The quantity of urban industrial land and transportation-water-conservation land will increase progressively, whereas the quantity of land and other construction land in rural residential areas will probably decrease; and 3) the scale and yearly average rate of construction land predicted with the improved array are smaller than that predicted with the original data.