Multi-factor nonlinear time-series ecological modelling for algae bloom forecasting

摘要

The current mechanism analysis method of algae bloom fails to take into consideration multi-factor nonlinear time-series characteristics of the ecological dynamics system in water, which leads to the low accuracy of algae bloom forecasting. In this paper, based on monitoring data of algae biomass (chlorophyll a concentration) and nutrient concentration, the nonlinear ecological dynamics model of algae bloom is constructed, which contains algae feeding and nutrient circulation, and model parameter optimization method is put forward by the combination of intelligent evolutionary algorithm and numerical algorithm. On this foundation, the effects of multiple factors time series such as illumination and temperature, which are the main influence factors of algae bloom, are considered into the algae bloom ecological system modelling. By using multi-factor time-series model to describe the variation of multiple influence factors, the algae multi-factor nonlinear time-series ecological dynamics model is constructed. A new method for algae bloom forecasting is put forward by multi-factor nonlinear time-series dynamic analysis. The example of Taihu Lake monitoring data shows that, compared with the current mechanism analysis method of algae bloom, multi-factor nonlinear time-series ecological dynamics model can better reflect dynamic characteristics of the algae bloom influence factors variation with time, and compared with the current forecasting methods, the forecasting results of algae bloom by the new method in this paper are more accurate.