基于贝叶斯网络的百色市生态环境综合评价与预测

Comprehensive Evaluation and Prediction of Ecological Environment in Baise based on Bayesian Network

为了研究百色市生态环境质量状况和识别关键影响因子,依据目的性、整体性、主导性、动态性和相关性等原则,建立百色市生态环境综合评价指标体系,利用AHP法、熵值法二者组合确定生态环境综合评价指标的权重,以驱动力-响应-效应为主线构建贝叶斯网络模拟和预测模型,通过敏感性分析揭示影响研究区生态环境的主要驱动因子,从定性和定量两个角度系统评估2013-2020年百色市生态环境质量状况及时序演变特征。研究表明,影响研究区生态环境的前五个核心驱动因子分别是森林覆盖率、年平均气温、年降水量、工业废水排放量、教育支出,百色市生态环境质量在2013-2020年不断得到改善,生态环境保护工作取得阶段性成效。最后,依据敏感性分析、人类活动因素对模型设置因果推理和诊断推理,对百色市生态环境综合模型进行预测研究,研究结果可为地方政府制定环境保护政策提供科学依据。

In order to study the quality of the ecological environment in Baise and identify key impact factors,this paper establishes the comprehensive evaluation index system of the ecological environment of Baise City according to the principles of purpose,integrity,dominance,dynamics and relevance,and uses the combination of AHP and entropy to determine the weight of the comprehensive evaluation index of the ecological environment,and constructs the Bayesian network simulation and prediction model with the driving force-response-effect as the main line,and through sensitivity analysis to reveal the main driving factors affecting the ecological environment in the study area,and the quality and temporal evolution characteristics of the ecological environment in Baise from 2013 to 2020 were systematically evaluated from both qualitative and quantitative perspectives.The study shows that the top five core driving factors affecting the ecological environment of the study area are forest coverage,annual average temperature,annual precipitation,industrial wastewater discharge and education expenditure.The ecological environment quality of Baise City has been continuously improved from 2013 to 2020,and the ecological environment protection has achieved phased results.Finally,based on sensitivity analysis and human activity factors,causal reasoning and diagnostic reasoning are set up for the model to predict the comprehensive model of the ecological environment in Baise.The research results can provide scientific basis for the local government to formulate environmental protection policies.