Study On Algae Bloom Monitoring Algorithm In Lake Taihu Based On CCD Data

In recent years, with the rapid development of industry and agriculture, the industrial,agricultural and living sewage containing a large number of nitrogen, phosphorus and other elements is discharged into the lake, making the lake serious eutrophication phenomenon,especially in the Yangtze River downstream in the vicinity of the lake. For example, Lake Taihu, Chaohu and others. There will be a serious outbreak of cyanobacterial blooms during every autumn festival. Cyanobacteria bloom directly affects the health of human, the development of economic and the balance of ecological. In this paper, we use our satellite of environment and disaster monitoring and forecasting CCD data. Through the method of Automatic Scattergram-Controlled Regression(ASCR), we do relative radiometric correction for the different imaging CCD data. And then we put forward a high precision extraction and easy operation algorithm that combining the Nomalized Difference Vegetation Index(NDVI)and Algae Pixel-growing Algorithm(APA) to extract the cyanobacteria bloom. And we also apply the algorithm in the satellite GF-1 that launched by China recently and satellite Landsat of United States. According to this study, the following conclusions can be drawn:1.After doing relative radiometric correction for the different imaging CCD data by the method of Automatic Scattergram-Controlled Regression, the correction results are ideal.Making the relative stability of the same objects in different images radiation values are the same, so we can monitor the cyanobacteria bloom in Lake Taihu through the different radiation in different images.2.We use the Nomalized Difference Vegetation Index to extract cyanobacteria bloom from each scene image in Lake Taihu area. Following we use the slope analysis method to determine the threshold of each scene image. By using statistical analysis method we finally determine a unified threshold to extract cyanobacteria bloom, helping solving the problem of different scene images with a different threshold that hard to large-scale batch processing.3.Through Algae Pixel-growing Algorithm to do pixel linear decomposition, theextraction accuracy can reach sub-pixel level. It provide a more accurate statistics of the area and distribution of algal bloom in Lake Taihu.4.By continuously monitoring algal bloom in Lake Taihu for a long time series from2009 to 2014, we can find that the year of 2013-2014 algal bloom area in Lake Taihu is reduction, smaller than the previous years, the water quality has been controlled and improved. The research also shows that the algorithm of detecting cyanobacteria bloom has a stronger recognition ability, a higher degree of automation and a higher extraction accuracy and can be used as the operation algorithm.5. Contrasting the satellite GF-1 that launched by China recently and satellite Landsat of United States, we find that they have a high correlation with our satellite of environment and disaster monitoring and forecasting CCD data. Thus, based on our satellite of environment and disaster monitoring and forecasting CCD data, we apply the algorithm in satellite GF-1data, as well as the Landsat series data to do relative radiometric correction, finally achieving the use of multi satellite platform for longer periods of time series and more high time resolution of cyanobacteria bloom real-time dynamic monitoring in Lake Taihu.