基于贝叶斯网络的渤海湾水体富营养化模型  

Water Eutrophication Model of Bohai Bay Based on Bayesian Networks

作　　者：袁德奎[1] 姚鹏辉[1] 徐晓甫[1,2] 聂红涛[3] 

机构地区：[1]天津大学机械工程学院,天津300072 [2]天津渤海水产研究所,天津300457 [3]天津大学环境科学与工程学院,天津300072

出　　处：《天津大学学报（自然科学与工程技术版）》2016年第3期320-325,共6页Journal of Tianjin University：Science and Technology

基　　金：教育部新世纪优秀人才支持计划资助项目(NCET-12-0406);天津市水产局青年科技资助项目(J2014-05)

摘　　要：基于渤海湾的现场监测数据,结合已有水体富营养化建模经验,用结构方程模型(SEM)对渤海湾水体富营养化因子与叶绿素a之间的因果关系进行了识别和验证.根据SEM提供的因果关系,建立了基于贝叶斯网络(BN)模型的渤海湾赤潮监控区海域水体的富营养化模型,并用监测数据对BN模型进行了检验和性能评价.研究结果表明:与物理因子相比,营养物质对渤海湾海域富营养化的影响较大;近年来无机磷一直为渤海湾浮游植物生长的限制因子,但是硅酸盐对浮游植物生长的影响程度越来越大.Based on the monitoring data and existing modeling experience in water eutrophication model, a structural equation model （SEM）was developed to understand the causality between eutrophication factors and chlorophyll a. According to the causal relationship between the variables provided by SEM, a water eutrophication model was built for the red tide-monitoring area in Bohai Bay based on Bayesian networks （BN）, and BN model performance was tested and assessed using the monitoring data. The results demonstrate that in Bohai Bay nutrients have greater influ- ence on the eutrophication than physical factors. In recent years, inorganic phosphorus has become the restriction factor influencing the growth of phytoplankton in Bohai Bay. However, the effect of silicate on phytoplankton growth is becoming more and more significant.

关 键 词：富营养化模型 贝叶斯网络 结构方程模型 叶绿素A 渤海湾 

分 类 号：X171[环境科学与工程—环境科学] TP181[自动化与计算机技术—控制理论与控制工程]