摘要
The Minqin oasis is surrounded on three sides by the Tengger Desert and the Badanjilin Desert, and it prevents these two deserts from converging. However, in recent years it has become the worst ecological environment in the Lake area due to deficient water resources, continual declines in the groundwater level and quality （increasing mineralization and salination）, which are causing in- creasing desertification. In this study, Landsat Thematic Mapper （TM） remote images from 1992, 1998, 2002, and 2006 of the Lake area of the Minqin oasis are interpreted to analyze the desertification evolution. A combination of an ArcObjects module and a cellular automata model is used to build a model simulating the desertification dynamics; the forecasting accuracy of this model is shown to reach up to 90%. The desertification situation in 2012 is forecasted by this model, and the results showed that, from 2006 to 2012, the green land area will be reduced by 999.92 hm2 （l.59 percent of the total oasis area）, the desertification land area will be reduced by 3,000.68 hrn2 （4.78 percent of the total oasis area）, and sand land area will increase by 4,000.6 hm2 （6.37 per- cent of the total oasis area）. The sand land is predicted to become more widespread, and more than 18% sand land will be distrib- uted in the center of green land in the Lake area. In other words, more and more abandoned green land （mined farm land） will be transformed into sand land, and this will intensify the desertification.
The Minqin oasis is surrounded on three sides by the Tengger Desert and the Badanjilin Desert, and it prevents these two deserts from converging. However, in recent years it has become the worst ecological environment in the Lake area due to deficient water resources, continual declines in the groundwater level and quality （increasing mineralization and salination）, which are causing in- creasing desertification. In this study, Landsat Thematic Mapper （TM） remote images from 1992, 1998, 2002, and 2006 of the Lake area of the Minqin oasis are interpreted to analyze the desertification evolution. A combination of an ArcObjects module and a cellular automata model is used to build a model simulating the desertification dynamics; the forecasting accuracy of this model is shown to reach up to 90%. The desertification situation in 2012 is forecasted by this model, and the results showed that, from 2006 to 2012, the green land area will be reduced by 999.92 hm2 （l.59 percent of the total oasis area）, the desertification land area will be reduced by 3,000.68 hrn2 （4.78 percent of the total oasis area）, and sand land area will increase by 4,000.6 hm2 （6.37 per- cent of the total oasis area）. The sand land is predicted to become more widespread, and more than 18% sand land will be distrib- uted in the center of green land in the Lake area. In other words, more and more abandoned green land （mined farm land） will be transformed into sand land, and this will intensify the desertification.
基金
supported by the National Natural Science Foundation of China (No. 40501073)
the Fundamental Research Funds for the Central Universities (Nos. 11CX05015A and 10CX04047A)
