Continuous Online Monitoring Of Eutrophication And Early Warning Of Cyanobacteria Blooms In Chaohu Lake

In recent decades, with the rapid development of industrial and agricultural production, the pollution of water eutrophication and cyanobacteria blooms has become one of the most serious water environment problems in the world wide. In China, the phenomenon of lake eutrophication and cyanobacteria blooms is particularly serious. Water blooms happen in most lakes, reservoirs and even rivers for different degrees. Large inland lakes like Taihu Lake, Dianchi Lake and Chaohu Lake have serious eutrophication. Usually, there is a large outbreak of cyanobacteria blooms in summer and autumn of each year.Cyanobacteria blooms which happen frequently in a large area for long time not only undermine the normal function of lake water and the environment, resulting in significant loss of fisheries production, but also threat to drinking water seriously, being harm to humans.And it has become an important environmental factor restricting economic and social sustainable development. At the present stage, there is no effective method for controlling cyanobacteria blooms from occurring. Therefore, real-time monitoring of changes in eutrophication status, timely occurrence of algal blooms in early warning, and taking emergency measures to reduce danger of algal bloom have great practical significance.In this paper, aiming at evaluation of entrophication and cyanobacteria blooms for Chaohu Lake, we selected the West-half of Chaohu Lake for studying. We also evaluated its degree of eutrophication using the integrated nutrition state index and carried out continuous online monitoring of eutrophication. Combining with real-time monitoring of water quality parameters and related meteorological parameters, we selected time series lag sensitive factor for cyanobacteria blooms prediction with Granger causality test, and established early-warning method of cyanobacteria blooms based on the vector autoregressive model (VAR model) to predict outbreak of cyanobacteria in Chaohu Lake. The study contents and main conclusions of each part are as follows:(1) We determined five indicators including chlorophyll-a (chl-a), total phosphorus (TP), total nitrogen (TN), permanganate index (CODMn)and transparency (SD) of10samples in West-half of Chaohu Lake in second half of2009by laboratory analysis, and evaluated the status of lake eutrophication with integrated nutrition state index method. The results show that the sampling area water in West-half of Chaohu Lake has been reached moderate-severe eutrophication, and it has the tendency of becoming more serious. In addition, integrated nutrition state index has an obvious seasonal trend.(2) Combining with continuous online monitoring data of eutrophication in Chaohu Lake in September and October,2009, we studyed the correlation and variation trends between the water quality parameters and algae concentration. The results show that cyanobacteria are the dominant species of Chaohu in September and October,2009, and the season of summer and autumn is a sensitive period of breakout of cyanobacteria blooms.The concentration of algae, water temperature, pH, dissolved oxygen and other parameters has significant day-night variable effect during this period. Water temperature and turbidity have a significant effect on chlorophyll-a content.(3) Basing on the data from eutrophication continuous line monitoring base station of Chaohu Lake in September and October,2009, we selected a variety of time series lag sensitive factors for cyanobacteria bloom with Granger causality test method. Analysis showed that the lag-sensitive factors of chlorophyll-a prediction for1day prediction contain dissolved oxygen, temperature and turbidity; the lag-sensitive factors of chlorophyll-a prediction for3days prediction include conductivity, temperature, turbidity and atmospheric pressure.(4) We established early-warning method of cyanobacteria blooms in Chaohu Lake based on the VAR model, and fitted predicted value with actual value of chlorophyll-a for1day and3days. The results showed that this model has a good reliable predicting effect and it is fit for short-term forecasting.