Author(s): Xiaoqing Zhang; Qiuwen Chen; Ruonan Li; Fei Ye
Keywords: Cyanobacteria blooms; Cellular automata; Hybrid evolutionary algorithm; Thiessen polygons
Abstract: Cyanobacteria bloom is a serious problem around the world and has caused severe ecological and social-economic damages. Data-mining models have proven effective to predict the blooms; however, few of them can provide the spatial dynamics. Although physical-based numerical models are an alternative, they are mostly too complicated to be well calibrated and applied in practice. This study took Taihu Lake as the case and developed a model, which embedded data-mining technique in a cellular automata configuration, so as to obtain the spatial-temporal dynamics of cyanobacteria blooms. The lake is divided into 31 polygons according to the 31 monitoring stations, by using Voronoi method. The water quality and ecological data are collected from the monitoring stations during 2008~2011. For each stations, hybrid evolutionary algorithm (HEA) was applied to establish a predictive formula of algal biomass in relation to physical-chemical parameters, by using the collected data. The formulae served incorporating with local interactions between polygons as the evolution rules of cellular automata. The results showed that the developed model has very promising performance in capturing the spatial-temporal dynamics of algal biomass in the lake.