【作者】 刘彦花；

【导师】 吴湘滨；

【作者基本信息】 中南大学 ， 国土资源信息工程， 2010， 博士

【摘要】 随着社会经济的发展和人类文明的推进,地质灾害给人类生存发展带来的影响越来越不可忽视,人类已将其看成一种不可不防、不可不治的灾害。地质灾害风险评估作为自然灾害研究的一项重要的基础内容,对认识地质灾害程度,制定减灾规划,部署防治工程,提高灾害管理水平具有十分重要的意义。近年来,国内外学者在灾害评估方面开展了大量的研究工作,但由于灾害评估是一个新的领域,所涉及的内容广泛,不仅包括自然科学,而且包括社会科学,已有的研究远没有形成完善的科学体系,现有的应用水平仍不能满足社会经济发展的减灾需要。因此,在地质灾害评估方面仍需深入的研究和进一步的探索。梧州是广西典型的山城,城市依山傍水而建,随着城市规模的扩张和各类工程的兴建,频发的地质灾害对人民的生命和财产构成越来越严重的威胁。本文借助科学的新手段开展梧州市地质灾害成灾机制与风险评估研究,不仅在丰富地质灾害研究理论与方法方面进行了新的探索与尝试,也对梧州的地质灾害研究具有重要的理论和实际意义。具体内容包括：(1)在收集、整理梧州市基础地理数据、地质环境背景数据和地质灾害专题数据的基础上,总结了梧州市地质灾害数据的多源性特点,借助地理信息系统、遥感、全球定位系统,实现了梧州市地质灾害多源数据的一体化集成,结合空间数据库技术,构建了梧州市地质灾害空间数据库,为下一步地质灾害成灾机制分析和风险评估奠定了数据基础；(2)在梧州市地质灾害空间数据库的基础上,借助地学信息图谱理论,充分利用GIS空间统计分析、叠置分析和缓冲区分析等空间分析功能及DEM建模,详细分析了梧州市各类地质灾害时、空分布状况,系统地总结了梧州市地质灾害发育特征与形成条件,实现了地学信息图谱理论与GIS支持下的地质灾害分布规律与成灾机制分析,为下一步灾害评估指标体系的构建提供了更加科学的参考依据；(3)在分析空间数据挖掘理论与方法的基础上,探讨了信息量法、模糊综合评判法、人工神经网络等地质灾害评估中常用的空间数据挖掘方法,提出运用粗集理论与地学粗空间相结合的方法进行梧州市地质灾害评估,根据梧州市地质灾害成灾机制地学信息图谱分析结果,构建了梧州市地质灾害危险性与承灾体易损性评估指标体系；(4)运用粗集与地学粗空间理论,建立了梧州市地质灾害危险性与承灾体易损性评估知识表达系统与决策表,构建了模型实现的具体算法,实现了评估指标体系优化与权重确定,在此基础上,结合模糊综合评判原理对决策表进行危险性与易损性定量分值计算,划分出地质灾害危险性与承灾体易损性等级,并通过野外实地调查和信息量模型结果对比,验证了评估结果的正确性；(5)在地质灾害危险性与承灾体易损性评估结果的基础上,实现了梧州市地质灾害风险评估,将梧州市划分为地质灾害高风险、中风险、低风险和极低风险四个等级,各等级按空间位置不同划分为若干亚类,并利用GIS实现评估结果可视化与区划制图；(6)系统地总结了目前梧州市地质灾害防治存在的问题,初步提出了梧州市地质灾害防治的对策与建议。更多还原

【Abstract】 With the development of society economy and civilization, geologic hazards have so important effects that human have to provide prevention and cure. As a essential and basic content of natural hazards, geologic hazards risk assessment presents very prominent significance for recognizing, controlling and managing geologic hazards. In the past few years, scholars have launched abundant researches on hazards assessment. However, hazards assessment is a new field which involves wide contents including natural sciences and social sciences so that existing researches are not perfect, and the application level still can not meet demand of reducing hazards. More in-depth studies and further explorations are needed on geologic hazards assessment.Wuzhou is classic mountain city in Guangxi which is at the foot of hill and beside river. Along with city outspreading and project constructing, frequent geologic hazards cause increasing serious threat to people’s life and properties. So researches on geologic hazards mechanism and risk assessment by using new scientific approches are developed in the theme, it not only proceeds with a new attempt for enriching geologic hazards therories and approaches, but also has important academic and applied significance for researching geologic hazards in Wuzhou.(1) Through collecting, sorting out basic geographic data, geologic environment background data and geologic hazards topic data in Wuzhou, multiple source character of geologic hazards data is concluded. With the benefit of 3S, integration of geologic hazards multiple source data is realized. Combining with spatial database techonlogy, geologic hazards spatial database is structured. It lays data foundation for next geologic hazards mechanism and risk assessment;(2) On the basis of geologic hazards spatial database, by drawing support from geo-infomatic atlas, using GIS spatial statistical analysis, overlaying analysis, buffer analysis and DEM modelling, geologic hazards time and space distribution is closely analyzed, geologic hazards growth feature and formation conditons are concluded. Based on geo-informatic atlas and GIS, geologic hazards distribution rules and mechanism analyses are implemented. It provides sceientific reference frame for next constructing assessment indicator systems;(3) Through analysing SDM theories and means, discussing some general SDM means in geologic hazards assessment such as information content model, fuzzy comprehensive assessment and neural network, rough sets and geo-rough space are put forward to assess geologic hazards. In accordance with geo-informatic atlas analysis results of hazards mechanism, geologic hazard danger and element vulnerability assessment indicator systems are established;(4) Hazard and vulnerability assessment knowledge representation system and decision table are built up based on rough sets and geo-rough space. At the same time, concrete algorithms to accomplish model are constructured, so assessment indicator system is optimized and their weights are confirmed. Combining with fuzzy comprehensive assessment, indicator weighs and their value in decision table, gradations of hazard and vlunerability are divided. Moreover, by comparing with information content model and field surveying, the assessment results above-mentioned are certified as correct;(5) Based on assessment results of hazard and vulnerability, geologic hazards risk assessment is accomplished, it is divided into four gradations such as high risk, middle risk, low risk and very low risk, every gradation is also divided into several subtypes. With the aid of GIS, assessment results are visualized;(6) Problems on geologic hazards prevention and cure in Wuzhou are systematically generalized, and moreover, some countermeasures and proposals are gived.更多还原