节点文献
基于最小加速法的二维地表形变场误差分析及沉降风险综合评估
Error Analysis of 2-d Ground Deformation Field Based on the Minimum Acceleration Combination Technique and Research on Comprehensive Risk Assessment Method of Land Subsidence
【作者】 王强;
【导师】 赵卿;
【作者基本信息】 华东师范大学 , 地图学与地理信息系统, 2020, 硕士
【副题名】以金砖国家典型沿海城市为例
【摘要】 城市重大基础设施形变及洪涝灾害是金砖国家沿海城市面临的共性挑战。由填海造陆、地铁和高层建筑等基础设施修建导致的地面沉降,以及由风暴潮带来的城市内涝正严重威胁着沿海城市的公共安全。因此需要对城市及防洪工程进行形变监测和对地面沉降进行风险评估。本文以金砖国家典型特大沿海城市俄罗斯圣彼得堡和中国上海为研究区,主要进行了以下研究工作:(1)开发了基于时间向量集合改进的最小加速二维形变分解方法(Modified Minimum Acceleration,MMinA),并基于此方法获取了圣彼得堡城区和综合防洪工程(St.Petersburg Flood Prevention Facility Complex,FPFC)的二维时序形变场。研究结果表明圣彼得堡市的沉降主要分布在新成陆和城市地铁沿线区域,年平均沉降速率分别达到20 mm/year和12 mm/year。MMinA二维形变分解结果反映出该市的形变以垂直方向为主。在对FPFC进行稳定性分析时发现,FPFC的D3段同时出现了垂向和水平向的结构变形,这需要引起重视。(2)以失相干导致的噪声为源误差,借助误差传播理论工具,系统的构建了针对MMinA的误差分析方法。并在圣彼得堡的应用研究中发现MMinA垂向分解精度优于东西向,且随着研究目标相干系数的增大分解精度也会随之提升。(3)基于多源社会感知数据和高分辨率小基线集(Small Baseline Subset,SBAS)时序形变监测数据,应用模糊神经网络技术在公里格网尺度下建立了上海市地面沉降风险综合评估模型。模型结果表明,上海市地面沉降潜在中、高风险多发于新成陆区。在中心城区中,黄浦区相较其他区的沉降风险中高等级面积比例更大,而在郊区中闵行区的沉降风险中、高等级面积比例最小。上海市各沉降风险等级面积比例约为稳定(Ⅰ):低风险(Ⅱ):中风险(Ⅲ):高风险(Ⅳ)=65:26:6:3。
【Abstract】 Deformation of major urban infrastructures and urban waterlogging are common challenges affecting the typical coastal cities in BRICS countries.Land subsidence caused by land reclamation,subway and high-rise building construction,and urban flooding caused by storm surges are seriously threatening the public safety of coastal cities.Thus,it is necessary to monitor the deformation of cities and flood control projects and conduct a risk assessment of ground subsidence.In this paper,the typical coastal cities of BRICS countries,St.Petersburg,Russia,and Shanghai,China,are used as research areas,and this paper mainly conducts the following research works:(1)A Modified Minimum Acceleration Combination Technique(MMinA)is developed to obtain the two-dimensional time-series deformation field of St.Petersburg city and St.Petersburg Flood Prevention Facility Complex(FPFC).The results show that the subsidence of St.Petersburg is mainly distributed in land-reclaimation area and along the Metro Lines,with the mean subsidence velocities of 20 mm/year and 12 mm/year,respectively.The MMinA decomposition results shows that the deformation of the city is mostly vertical with respect to the east-west direction.Of great concern is the study of the stability of the FPFC that was built to prevent the city from extreme flood events.In particular,our analysis has revealed that the D3 section of the facility is more prone to deformation and that the lateral movements also affect the structure,which is of high relevance for urban planners and security practitioners.(2)The error distribution law of MMinA was found with the using of an error analysis method proposed in this work.Subsequently an error analysis of the MMinA decomposition results reveals that the accuracy of the Up-Down deformation components is better than that of the East-West deformation components and it improves as the average spatial coherence of the selected pixels increases.(3)A comprehensive risk assessment model of land subsidence was established with the using of multi-source social perception data,high-resolution Small BAseline Subset(SBAS)deformation time series and fuzzy neural network technology.The model results show that the potential medium and high risks of land subsidence in Shanghai are mostly located in land-reclaimation areas.In the central urban area,Huangpu District has a higher proportion of areas with medium and high-level land subsidence risks than other administrative districts.As for the suburbs,Minhang District has the smallest proportion of areas with medium and high-level land subsidence.And the model also shows that area ratio of each land subsidence risk level in Shanghai is approximately: stable(Ⅰ): low risk(II): medium risk(III): high risk(IV)= 65: 26: 6: 3.
【Key words】 Ground Deformation Decomposition; Error Analysis; Risk Assessment; Fuzzy Neural Network;