研究报告
马宇博,高广阔.基于节点重要性评价的京津冀雾霾污染网络研究[J].环境科学学报,2018,38(6):2287-2296
基于节点重要性评价的京津冀雾霾污染网络研究
- Research on haze pollution network of Beijing-Tianjin-Hebei region based on node importance evaluation
- 基金项目：国家社会科学基金资助项目（No.15BTJ017）
- 作者
- 单位
- 马宇博
- 上海理工大学管理学院, 上海 200093
- 高广阔
- 上海理工大学管理学院, 上海 200093
- 摘要：近几年来，雾霾污染已经发展成为一个区域性的环境问题.区域内城市间雾霾污染的相关性错综复杂，呈现出复杂网络的结构形态.从整个雾霾污染大网络中去除一些联系不紧密、重要性不高的城市节点，进而找出真正发挥关键作用的核心子网络，对于整个区域的联防联控都有着重要的现实意义.本文将京津冀地区13个城市作为节点，选取2013年10月28日至2017年10月17日，共1488 d内每个城市的日均PM2.5浓度数据作为样本.首先，根据城市间PM2.5日均浓度的相关系数作为是否连边的判别标准，并由相关系数与城市间距离的比值构造出边的权重，通过连边状况去除了张家口这个孤立节点.其次，通过计算节点的加权集群系数、网络效率与脆弱性和采用基于度和集群系数的节点重要性综合评价方法对节点的重要性进行评价.接着，综合3种节点重要性评价方法得到节点重要性排名，由高到低分别是保定、唐山、廊坊、石家庄、天津、北京、沧州、衡水、邢台、秦皇岛、邯郸、承德.并根据节点重要性评价结果，从整个京津冀区域的大网络中提取出联系最为紧密、相关性最强的一个雾霾污染子网络.最后，文章结合京津冀地区雾霾污染子网络的结构特点与现有大气污染防治的政策法规，提出了促进区域内雾霾污染防治一体化的3点建议.
- Abstract：In recent years, haze pollution has been a regional environmental problem. The correlation of haze pollution among cities in a region is quite complex, showing the structure of complex network. Removing some of nodes which are neither closely related nor important from the whole haze pollution network and then figuring out the core subnetwork that plays a key role in the network are of great practical importance to the joint prevention and control of the entire region. Therefore, it is very significant to evaluate the importance of nodes. In this paper, 13 cities in Beijing-Tianjin-Hebei region were selected as the nodes, and the daily average PM2.5 concentration data of each city in 1488 days from October 28, 2013 to October 17, 2017 were chosen as samples. First of all, a criterion of edge connection was set according to the correlation coefficient of PM2.5 concentration between two cities, the weight of edges was built through the correlation coefficient and the distance between cities, and the isolated node-Zhangjiakou was removed in terms of the edge connection. Secondly, the importance of nodes was evaluated by calculating the weighted clustering coefficients, network efficiency and vulnerability of nodes, and using the node importance evaluation method based on degree and clustering coefficient. Then, node importance ranking was obtained based on the three methods comprehensively, the rank from high to low was Baoding, Tangshan, Langfang, Shijiazhuang, Tianjin, Beijing, Cangzhou, Hengshui, Xingtai, Qinhuangdao, Handan and Chengde. Based on the results of node importance evaluation, a haze pollution subnetwork which is most closely related and most relevant was extracted from the whole network of Beijing-Tianjin-Hebei region. Finally, combining with the structural characteristics of the haze pollution subnetwork of Beijing-Tianjin-Hebei region and the existing policies and regulations on air pollution prevention, three suggestions were put forward to promote the integration of haze pollution prevention and control in the region.
摘要点击次数： 1097 全文下载次数： 1413
