Multi-factor Analysis of Algal Blooms in Gate-Controlled Urban Water Bodies by Data Mining.
Ke Li,Te Xu,Jinying Xi,Haifeng Jia,Zhengjuan Gao,Zhaoxia Sun,Dingkun Yin,Linyuan Leng
DOI: https://doi.org/10.1016/j.scitotenv.2020.141821
2021-01-01
Abstract:Intense human disturbance has made algal bloom a prominent environmental problem in gate-controlled urban water bodies. Urban water bodies present the characteristics of natural rivers and lakes simultaneously, whose algal blooms may manifest multi-factor interactions. Hence, effective regulation strategies require a multi-factor analysis to understand local blooming mechanisms. This study designed a holistic multi-factor analysis framework by integrating five data mining techniques. First, the Kolmogorov-Smirnov test was conducted to screen out the possible explanatory variables. Then, correlation analyses and principal component analyses were performed to identify variable collinearity and mutual causality, respectively. After collinearity and mutual causality were treated prudently by using orthogonalization and instrumental variables, multilinear regression can be properly conducted to quantify factor contributions to algae growth. Lastly, a decision tree was used innovatively to depict the limiting threshold curves of each driving factor that restricts algae growth under different circumstances. The driving factors, their contributions, and the limiting threshold curves compose the complete blooming mechanisms, thus providing a clear direction for the targeted regulation task. A typical case study was performed in Suzhou, a Chinese city with an intricate gate-controlled river network. Results confirmed that climatic factors (i.e., water temperature and solar radiation), hydrodynamic factors (i.e., flow velocity), nutrients (i.e., phosphorus and nitrogen), and external loadings contributed 49.3%, 21.7%, 21.3%, and 7.7%, respectively, to algae growth. These results indicate that a joint regulation strategy is urgently required. Future studies can focus on coupling the revealed mechanisms with an ecological model to provide a comprehensive toolkit for the optimization of an adaptive joint regulation plan under the background of global warming. (C) 2020 Elsevier B.V. All rights reserved.