Dynamic monitoring and prediction of Dianchi Lake cyanobacteria outbreaks in the context of rapid urbanization.

Water crises have been among the most serious environmental problems worldwide since the twenty-first century. A water crisis is marked by a severe shortage of water resources and deteriorating water quality. As an important component of water resources, lake water quality has deteriorated rapidly in the context of fast urbanization and climate change. This deterioration has altered the water ecosystem structure and influenced lake functionality. To curb these trends, various strategies and procedures have been used in many urban lakes. Among these procedures, accurate and responsive water environment monitoring is the basis of the forecasting and prevention of large-scale cyanobacteria outbreaks and improvement of water quality. To dynamically monitor and predict the outbreak of cyanobacteria in Dianchi Lake, in this study, wireless sensors networks (WSNs) and the geographic information system (GIS) are used to monitor water quality at the macro-scale and meso-scale. Historical, real-time water quality and weather condition data were collected, and a combination prediction model (adaptive grey model (AGM) and back propagation artificial neural network (BPANN)) was proposed. The correlation coefficient (R) of the simulation experiment reached 0.995. Moreover, we conducted an empirical experiment in Dianchi Lake, Yunnan, China using the proposed method. R was 0.93, and the predicting error was 4.77. The results of the experiment suggest that our model has good performance for water quality prediction and can forecast cyanobacteria outbreaks. This system provides responsive forewarning and data support for lake protection and pollution control.