Research on ANN-based Pre-warning Water Bloom Model of LiuHai Lake in Beijing

Aimed at the Problems of water bloom erupting often in the cities of our country, a Pre-Warning Model for water bloom is developed by using the ANN method; in order to adopt the emergency response measures and avoid erupting water bloom, and provide scientific basis for improving the water environment of river and lake system. The results of the research indicates: (1) water system dynamic condition is the main factor causing the water bloom, the change of water system dynamic condition is the best way for avoiding the eruption of water bloom. (2) The present concentration of Chl-a is main pre-warning index, if the water bloom has erupted, the measure of emergency allocation water may relax the tension of water bloom, so far as to reduce red into green pre-warning situation. (3) The velocity of flow is the decisive factor of water bloom eruption. When the velocity of flow is great enough the water bloom erupts seldom; and when the velocity is zero, even if the concentration is not very big, the water bloom will erupt possibly after 7 days. (4) The temperature of water is not the essential condition of water bloom. Even if the temperature is very lower, when the velocity of flow is very slow and the nutrition level is great enough, the water bloom will erupt possibly. (5) The eutrophication is not the inevitable factor. Under the good water dynamic condition, even if the nutrition level is very high, the water bloom will not erupt. Only when the temperature is suitable and velocity of flow is slow, the water bloom will erupt.