An Accurate Algorithm To Estimate Cyanobacteria Bloom Area In Lake Taihu Using MODIS Images And Its Applications

Eutrophication is one of the most serious problems in lakes. Taihu is a typical eutrophic lake in China, with frequent algal blooms. Compared with traditional monitor methods, remote sensing technique is more suitable for monitoring algal blooms dynamically for a long time. Among different images, MODIS images are the optimal data for monitoring algal blooms in Lake Taihu because of free of charge, high temporal resolution and abundant spectral information, and have been widely applied for monitoring algal blooms, as a part of daily work of management agencies. However, different methods, threshold value and large degree of subjectivity have resulted in incompatible estimates. In addition, the current methods neglect the coverage of algal mats, causing the inaccurate area result.In order to solve these problems, based on FAI detecting methods, we propose a more scientific and reasonable method to compute algal bloom area in this paper, that is Algae Pixel-growing Algorithm(APA). This algorithm identifies the "seed" pixles and defines the seed pixcels’ coverage, then the algae coverage at each pixels can be determined iteratively by establishing the linear relation with nearby pixcels. The APA results are validated using concurrent higher-resolution TM/ETM+ observations. The RMSE and RE between MODIS estimates and TM/ETM+ estimates are 15.2 km2 and 9.9%, showing accurate performance. The APA method may serve as an objective method to determine the bloom severity in both near real-time and history, leading to more accurate information to help make management decisions.In this study, the algal bloom coverage of Lake Taihu from 2001 to 2013 have been identified using APA method. Then we analyse the spatial and temporal distribution of algal bloom in Lake Taihu, and figure out the relationship between algal bloom area and meteorological, nutrition factors. The results indicate that: Before 2004, the algal bloom occurred mainly in Zhushan Bay, Meiliang Bay. There were no obvious algal bloom in Central Lake and South Lake almostly. Gong Bay and West Lake occurred obvious algal blooms only in summer. After 2004, Gong Bay and West Lake occurred more high intensity algal blooms, and the algal blooms spreaded to Central Lake and South Lake. The algal blooms in Taihu Lake occurred earlier and showed longer durations than 2001～2004, especially during 2006-2008. Since 2008, the algal bloom problem showed an upward turn. High intensity algal blooms occurred less, and low intensity blooms dominated. But the algal bloom problem were still serious than before 2004.The nitrogen and phosphorus in Lake Taihu are at high concentration, and phosphorus shows no obvious correlation with algal bloom area. Meanwhile, the analysis result exhibites a negative correlation between nitrogen and algal bloom area. By analyzing the relation between algal bloom area and meteorological factors, we find that wind speed less than 3m/s, at NNW、NW、ESE, and 20-35℃ temperature are favorable for the occurrence of algal bloom in Lake Taihu.