干旱区生态环境知识资源中心

Nowadays, the study about the mutative characters of wetland landscape mainly focus on pattern change. They are few people to take research on the wetland landscape evolvement on the basis of hydrological variation. The study area of Yinchuan is located in the region of and and semi-arid areas, northwest China. And this paper presented a new method for the restoration of the wetland landscape areas in the and and semi-arid region for 50 years. Taking the precipitation as variation, the relationship between the precipitation and wetland area was established via the Markov Chain Model Method, with the unit precipitation as step with the support of RS and GIS technology. Because the precipitation has lag effect on the wetland evolution between the inter-years, therefore, we adopted three-year-average slippage disposal of the precipitation data to carry on the reconstruction of wetland restoration between different periods. The simulated data were verified through the areas of wetland landscapes extracted from the remote sensing images of 1991, 1999 and 2006. The accuracies of these three years' simulated data reached 93.8245%, 91.1791% and 91.8702% respectively. The total forecasting accuracy reached 92.1181%. So, the simulative method is feasible and the simulated results are creditable.