节点文献

三峡库区滑坡灾害风险评估研究

Landslide Risk Assessment in the Three Gorges Reservoir

【作者】 彭令

【导师】 黄波; 牛瑞卿;

【作者基本信息】 中国地质大学 , 地球探测与信息技术, 2013, 博士

【摘要】 自二十世纪初中期以来,随着社会经济的大规模发展,人类活动空间范围的逐渐扩展,以及重大工程活动对地质环境扰动程度的不断加剧,加之受到全球极端气候变化等因素的影响,滑坡事件发生的频率和强度均呈增长之势,所造成的人员伤亡和经济损失也不断加大。长江三峡库区地质条件复杂,人类工程活动强烈,滑坡等地质灾害广泛发育且频繁发生,对三峡水库调度运营及当地人民生命财产安全构成严重危害,直接或间接影响该区域经济发展和社会安定。因此,研究三峡库区滑坡灾害风险,以及预测未来气候、人类工程活动变化条件下滑坡灾害风险的变化情况及趋势,是实现库区防灾减灾战略及保障水库正常运营的重大而迫切需求,具有十分重要的科学和现实意义。论文以区域滑坡灾害风险评估的理论方法为基础,以地理信息系统、遥感、概率统计及数据挖掘方法为主要技术手段,以实现三峡库首区(秭归至巴东段)滑坡灾害风险评估为目的。在野外调查、相关数据资料收集及已有研究成果的基础上,从区域滑坡发育特征和典型滑坡专业监测变形特征分析入手,预测滑坡发生的空间位置、时间频率和强度以实现滑坡危险性评价。以承灾体易损性定义为出发点,建立灾害强度和抗灾能力评价指标体系并定义其计算公式,以此构建承灾体易损性评价模型。融合经济学方法核算承灾体价值,采用定量风险模型估算滑坡灾害生命和经济风险,并建立风险标准对风险分析结果进行评价。通过分析降雨和土地利用变化规律,预测中长期滑坡危险性及灾害风险的变化情况及趋势,最终实现三峡库首区滑坡灾害风险评估,取得的主要成果及结论如下:(1)提出了三峡库首区滑坡空间预测评价指标体系的粗糙集分析方法,在此基础上建立了遗传—支持向量机滑坡易发性预测耦合模型以及滑坡强度分析概率模型。分析结果表明孕育和影响土质滑坡的关键环境因素包括高程、坡度、坡向等16个因子,而岩质滑坡则包含高程、坡度、斜坡形态等17个因子,表明不同滑坡类型既受相同环境因素作用但又有各自不同的环境因素影响。滑坡易发性预测结果显示高易发区主要分布在长江干流的香溪至巴东段,以及支流归州河、童庄河、青干河沿岸和香溪河左岸。野外验证及模型质量评估表明预测结果与区域构造背景、工程地质条件、地形地貌等实际调查情况吻合,模型预测结果可靠,其中不确定性主要集中在中等易发区,而低易发区和高易发区的不确定性较小,土质和岩质滑坡的预测精度分别为86%和85%。滑坡强度分析结果显示发生5-6级强度的土质和岩质滑坡概率分别为0.48和0.35。滑坡危险性预测结果表明在多种构造行迹复合部位和新构造活动相对强烈地区,含有软弱面(带)的层状碎屑岩和有软弱基座的层状碳酸盐岩地层,以及河流顺褶皱轴向发育形成的顺向岸坡和人类工程活动强烈区是滑坡高危险性地段,例如新滩附近、树坪至范家坪、东壤口附近、巴东县城附近和归州河、童庄河、青干河沿岸及香溪河左岸,其滑坡发育具有空间密、频率高和强度大等特点。(2)建立了三峡库首区滑坡灾害的承灾体易损性分析模型,提出基于此模型的滑坡灾害风险评估方法。易损性分析结果表明三峡库首区内的人口、建筑物、生命线工程和土地资源等承灾体的易损性高,并且高易损性的承灾体主要集中在滑坡高危险性区内。滑坡灾害生命高风险区主要分布在巴东县城、东壤口、郭家坝等集镇和建有学校、企事业单位等人口密集的地区:经济高风险区主要集中在城镇和交通建设用地等经济价值大的地区。风险评价结果表明不可接受风险区主要分布在城镇和人口密集的村庄及经济价值大的建设用地区域,野外调查与验证表明滑坡灾害风险评价结果与实地灾害情况基本吻合。(3)预测了三峡库首区中长期降雨和土地利用变化条件下滑坡危险性和灾害风险的变化情况及趋势。预测结果显示21世纪前期该区年降雨量有所减少,而中后期则将增大。土地利用中建设用地和林地将持续性增长,耕地前期增多而后期有所减少,灌木林地持续性减少,水体基本维持稳定,并有少量增加。滑坡危险性整体上有增大趋势,而局部地区将有所降低,并且前期危险性变化主要集中在巴东地区,而后期则转移至秭归地区。滑坡灾害经济风险在2010至2050年将持续性增长,并且前二十年经济风险增长相对较大:生命风险在2010到2030年表现出增长态势,而后二十年则呈现出一定程度的降低。

【Abstract】 Since the early and middle2th century, with the large-scale development of social economy, the space scope of human activities expands gradually. And the disturbance degree of major engineering activities to geological environment is deepening constantly. Combined with the influence of the global extreme climate change, the frequency and strength of occurred landslide is in a growing trend. And the caused casualties and economic losses are increasing continually. The complex geological conditions of the Yangtze river Three Gorges Reservoir, the severe human engineering activities, widely development and frequently happened geological disasters of landslide have posed a serious hazard for the Three Gorges Reservoir operation scheduling and the local people’s life and property security, and affect the regional economic development and social stability directly or indirectly. So, to study the Three Gorges Reservoir area landslide risk and to predict the future risk situation and change tendency under the condition of climate changes and human engineering activities are the great and urgent demand of the implementation of disaster prevention and mitigation strategy and to ensure the normal operation of the reservoir. It is of great importance to scientific and realistic significance.The thesis takes the regional landslide risk assessment theory as the foundation, and adopts geographic information system, remote sensing, probability statistics and spatial data mining method as the main technical means, to achieve the purpose of the Three Gorges Reservoir head region (Zigui-Badong segment) landslide risk assessment. On the basis of field investigation, relevant data collection and existing research results, from the perspective of the regional distribution characteristics and typical landslide professional monitoring deformation characteristics analysis, the thesis predicts the landslide happening spatial location, temporal frequency and landslide magnitude to achieve the landslide hazard assessment. It takes vulnerability of elements at risk definition as the starting point, extracting hazard intensity and resistance ability evaluation index system and defining its calculation formula, so as to establish a vulnerability assessment model of elements at risk. It uses economics accounting method to calculate the value of elements at risk, using the quantitative risk model to estimate landslide life risk and economic risk, and to establish risk criteria and evaluate the risk result. Through the analysis of rainfall and land use change rule, the thesis forecasts the landslide hazard and risk change and trend under the condition of future changes, realizing the Three Gorges Reservior head region landslide risk assessment. The main conclusions and results are as follows:(1) The thesis establishes the Rough Sets analysis method of the Three Gorges area landslide spatial prediction and evaluation index system, the Genetic Algorithm-Support Vector Machine susceptibility prediction coupled model and lanslide magnitude analysis model. The results show that the key environmental factors which gestate and affect soil landslide include elevation, slope angle, slope aspect, etc.16factors. And rock landslide consists of17factors, such as elevation, slope angle, morphology of slope, and so on. It proves that different landslide types are affected both by the same environmental factors and different environmental factors. Landslide susceptibility prediction results show that high susceptibility prone are mainly distributed in the Yangtze river from Xiangxi to Badong, and tributary Guizhou river, Tongzhuang river, Qinggan River along the bank and the left bank of Xiangxi river. The field validation and model quality evaluation show that the predicted results are matched with the regional tectonic setting, engineering geological conditions, topography and field investigation situations, etc.. The model prediction results are reliable, and the uncertainty mainly concentrates in the medium susceptibility zones. And the uncertainty of low susceptibility prone and high susceptibility prone is small. The prediction accuracy of soil landslide and rock landslide are86%and85%respectively. Lanslide magnitude analysis results show that the probability of5-6magnitude soil landslide and rock landslide are respectively0.48and0.35. Landslide hazard prediction results show that in the varieties of structure trackway composite parts and new tectonic movement relatively strong area, the areas which contain weak surface (belt) of layered clastic rock and have a weak foundation layer of carbonate rock, as well as rivers along the fold axial development formation of the dip slope and strong human engineering activities are high hazard in landslide. For example, around the Xitan, Shuping to Fanjiapin, around Dongrangkou, near Badong town and Guizhou river, Tongzhuang river, Qinggan River along the bank and the left bank of Xiangxi river, their landslide development have the characteristics of spatial density, high frequency and magnitude, etc..(2) The thesis establishes the vulnerability analysis model of the Three Gorges Reservior head region landslide hazard. The vulnerability analysis results show that the population, buildings, lifeline engineering and land resources in the Three Gorges Reservior head region are high vulnerability. And the high vulnerability of elements at risk mainly focuses in the high hazard landslide area. The life high risk areas mainly distribute in the towns of Badong County, Dongxiangkou, Guojiaba and so on, and also in the densely populated areas such as schools, enterprises and institutions. The economic high risk areas mainly concentrate in areas of the large economic value:the urban and traffic construction land etc.. The risk assessment results show that the unacceptable risk area mainly distributes in towns and villages with a dense population and large economic value of construction land area. The field survey and verification show that the landslide risk assessment results are basically matched with the field disasters. (3) The thesis predicts the Three Gorges Reservior head region mid-and-long term landslide hazard and risk change and trend under the condition of rainfall and land use changes. The prediction results show that the annual rainfall will decrease in the early of the21st century and will increase in the middle and later period. Construction land and forest land will continue to grow. Cultivated land will first increase and then decrease. Shrub land will decrease continuously. The water area will remain stable basically, but there is a small amount of increase. Landslide hazard will be in an increasing trend on the whole, but part of the areas will reduce. And the early hazard changes will mainly concentrate in Badong area, later will transfer to Zigui region. The landslide economic risk will grow sustainably from2010to2050. And in the first twenty years, the economic risk growth is relatively large. The life risk is in a grow trend from2010to2030, and in a degree of drop in the later twenty years.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络