Study On The Dynamic Changes Of Desert Wetlands In Inner Mongolia

Using CBERS satellites data, as well as images, selected the larger area, clearer image of the lakes in the Ulan Buh Desert, Badain Jaran desert, Hobq desert and Tengger desert, through the ERDAS and ArcGIS software processing of remote sensing images. Take for wetlands in the four deserts as the research object, has lakes area dynamic testing. Combination of meteorological data in the study area, were carried out lake area and temperature, precipitation, evaporation and relative humidity correlation analyzed. As the Ulan Buh Desert for the typical study area, used of remote sensing software to calculate the normalized difference vegetation index NDVI, combined with vegetation coverage of field investigation, has been studied correlation between the both. Established of vegetation index based on vegetation coverage regression model, and do the model verification.The main results are as follows:(1)Desert wetlands of Inner Mongolia belongs to continent type and closed-type lakes, Ulan Buh desert lakes showed an early atrophy, post-expansion; lakes in Badain Jaran desert and Tengger desert area was not changed much; Hobq desert lakes was atrophied year by year. Four desert lakes area of expansion and contraction recorded weather data with the corresponding period the average annual temperature, annual rainfall, average annual evaporation, annual changes in relative humidity has a good consistency, reflecting the CBERS images in lake waters change detection has a higher availability.(2)Normalized difference vegetation index and vegetation coverage of the correlation coefficient has been reached 0.888, correlation degree was higher; the established of vegetation index based on vegetation coverage regression model:Y = 24.738 +107.319 NDVI. Experience certificated, fitting precision was attained 92%, indicating the establishment of a NDVI-based linear regression model inversion of vegetation coverage and the measured values very closed. Based on regression model calculating vegetation coverage, studied timing sequence dynamic changes of vegetation coverage of desert lake 1and2, the results showed vegetation coverage has no significant change law, and no correlation with lake area. Therefore, in terms of remote sensing images by based on the NDVI simple linear regression model to monitor the vegetation coverage in Inner Mongolian desert wetlands was simple and effective.