网格化城市内涝风险积水深度CID计算模型  

CID calculation model for grid urban waterlogging risk depth

作　　者：薛丰昌[1,2] 杨盟 陈剑飞 信佳岑 XUE Fengchang;YANG Meng;CHEN Jianfei;XIN Jiacen(Meteorological Disaster Geographic Information Engineering Laboratory,Nanjing University of Information Science&Technology,Nanjing 210044,China;Hebei Key Laboratory of Resource and Environmental Disaster Mechanism and Risk Monitoring,Sanhe 065201,Hebei,China;Guangxi Zhuang Autonomous Region Lightning Protection Center,Nanning 530001,China)

机构地区：[1]南京信息工程大学气象灾害地理信息工程实验室,南京210044 [2]河北省资源环境灾变机理及风险监控重点实验室,河北三河065201 [3]广西壮族自治区防雷中心,南宁530001

出　　处：《安全与环境学报》2024年第11期4383-4390,共8页Journal of Safety and Environment

基　　金：河北省资源环境灾变机理及风险监控重点实验室开放基金项目(FZ248101);广西重点研发计划项目(桂科AB22080101);中国气象局流域强降水重点开放实验室基金项目(2023BHR-Y29)。

摘　　要：针对传统城市内涝风险评估方法存在较强主观性问题,将地理信息系统技术与云化信息扩散(Cloud-based Information Diffusion,CID)模型相结合,提出了一种网格化城市内涝风险积水深度CID计算模型。将郑州市中原区、二七区、金水区和管城回族区交界处街区作为研究区,对研究区2013—2021年气象站点逐小时降雨数据进行统计分析,提取24 h累计连续降雨量超40 mm的27场典型降雨过程作为降水过程样本,利用基于暴雨管理模型二次开发的城市雨洪模型构建研究区城市内涝模拟模型,对模拟模型进行参数率定、模型精确性验证后,通过选定的27场降水过程数据模拟形成研究区27幅百米网格内涝分布图,形成网格化最大雨强-最大积水深度二维数据集;在二维正态信息扩散模型中引入云化信息模型的期望E_(x)、熵E_(n)和超熵H_(e)指标,形成二维云化信息扩散模型;建立信息扩散模型知识表达系统,并用于表达信息扩散模型的信息量分配关系,实现网格单元风险积水深度的预测计算。研究区进行格网剖分后,共形成了5760个网格,以100 mm/h雨强为降水场景示例,计算研究区所有网格单元在示例降水条件下的风险积水深度预测值,形成区域内涝预报图。研究结果显示,该方法能够显著降低灾害样本信息中的模糊性和随机性,有效实现网格化内涝风险积水深度预测,有助于提高内涝防灾减灾技术能力。Traditional methods for assessing inland inundation risk in urban areas are often subjective and inaccurate in predicting disaster risks.To address this issue,we integrate Geographic Information System(GIS)technology with the Cloud-based Information Diffusion(CID)model to propose a grid-based calculation model for assessing water depth related to urban inland inundation risk.We selected the blocks at the intersection of Zhongyuan District,Erqi District,Jinshui District,and Guancheng Hui District in Zhengzhou City as our study area.We conducted a statistical analysis of hourly rainfall data from meteorological stations collected between 2013 and 2021 within this region.We selected 27 typical rainfall events,each with cumulative continuous rainfall exceeding 40 millimeters within 24 hours,as samples for our precipitation processes.A city rainfall flood model,based on the Storm Water Management Model(SWMM)and modified through secondary development,was employed to construct a simulation model for urban inland inundation in the study area.Following parameter calibration and model accuracy verification of the simulation model,we utilized the selected 27 precipitation events to simulate and generate inundation distribution maps at a hundred-meter grid scale in the study area.This resulted in a two-dimensional dataset representing maximum rainfall intensity and maximum water depth.We incorporated the expectation E_(x),entropy E_(n),and hyperentropy H_(e)indicators from the cloud information model into the two-dimensional normal information diffusion model,resulting in the development of a two-dimensional cloud information diffusion model.This established a knowledge representation system for the information diffusion model,enabling the expression of the information distribution relationship and facilitating the prediction and calculation of risk water depth in grid units.After subdividing the study area into grids,we created a total of 5760 individual grids.Using a rainfall intensity of 100 mm/h as an example,we calculat

关 键 词：公共安全 洪灾风险 积水深度 信息扩散理论 

分 类 号：X43[环境科学与工程—灾害防治]