Analysis Of Spatial Pattern Change Of Organic Pollution To Water Bodies Based On Landsat TM/ETM+

Huangshui river valley is the political, economic and cultural center of Qinghai province. In recently years, with the improvement in urban and industry, a great deal of unstandardized industrial waste and life sewage have been discharging, which make the pollution of the water more serious, especially the organic pollution, it has become the importance chemical substance which effects human healthy. Huangshui mainstream ranging form Xining to Ledu is typical, that it is the most densely populated and the most centralized town region, and at the same time that it is also the region where water resource is centrally used and water pollution is the most serious. Analyzing the spacial pattern change of the water organic pollution using remote sensing technology is valuable to protecting water resource and controlling water pollution.In this paper, a method and process which deal with the TM/ETM+ image is put forward. The main processes include relative radiative correction of atmospheric attenuation, normalizing the multi-temporal remote sensing images using pseudo-invariant features which are chosen with artificial method and extracting the water information using spectrum connection and threshold value that the model isIn this paper, the DO,COD,BOD and NH3-H are chosen as representative element estimating water quality. The single indicator method is used to appraise and analyze the water quality in the Huangshui mainstream ranging from Xining to Ledou The water quality is divided into four grades including clean water, light-polluted water, medium-polluted water and heavily-polluted, and the four grades are respectively given blue,green, yellow and red.TM1,TM2 and TM3 are chosen as the best band combination by analyzing the spectral feature of elements and the three bands are used to set up RS information model.Expert classification system is tentatively used in the construction of the RS information model. The better values which are used to divide the organic pollution level are identified by an amount of experimentation and analysis. Compared the two models that one is constructed by expert classification system and the other is done by density slice. The result shows that the former is better than the latter. Eventually the expert classification is chosen. The spatial distribution pattern of the water organic pollution is generated in the IMAGINE9.1 environment and the organic pollution levels are quantified. The research result shows that clean water is very little and most of water is polluted ranging from Xining to Ledou; the water above 95% is polluted in different level and the proportion of the light-polluted is the largest. Because other pollution types such as heavy metal pollution aren't considered, but there are a lot of industries in study area and the heavy metal pollution is very serious, the result is different with"water resources bulletin".