节点文献

基于多信息源的台风灾害实时评估系统研究

Reseaching on Dynamic Assessment of Typhoon Disaster System Based on Multi-sources

【作者】 刘少军

【导师】 何政伟;

【作者基本信息】 成都理工大学 , 地球探测与信息技术, 2011, 博士

【副题名】以海南岛为例

【摘要】 人类的生存与发展是和天灾人祸的侵害相伴随的。据世界气象组织有关专家统计,威胁人类生存的10大自然灾害有台风、地震、洪水、雷暴、龙卷风、雪暴、雪崩、火山爆发、热浪、山体滑波(泥石流)、海潮(海啸)等,在这10种自然灾害中台风是造成死亡人数之冠。由于灾害评估是全面反映灾情,进行减灾决策的基本依据,也是制定社会经济发展计划的重要参考资料,因此有必要开展台风灾害的损失评估。在台风灾害损失评估中,评估主要针对直接经济损失评估,常规的方法是通过计算台风灾害直接造成的各类动产和不动产修理或重置成本,一般要在灾后数天、数周甚至数月内计算出来,缺乏实时性。灾害损失是由众多灾害影响因素相互作用的结果,而这些影响因素中有些可以用精确的数学模型来度量,有些则无法用精确的数学模型来描述。同时台风是一个动态变化的复杂系统,其结构、路径、强度、风雨分布等在不同阶段会产生一系列急剧变化,在不同时刻对同一地点造成的影响也是动态变化的。在台风灾害的防灾减灾过程中,必须迅速掌握不同时间不同范围内灾害的损失情况,传统的统计方法在对台风灾害的应急、防灾减灾工作的实效性达不到,因此必须探索新的灾害评估方法。针对台风灾害评估的存在的一些问题,提出了以数值天气预报、自动气象站、多普勒天气雷达为主要信息源,以可拓方法为理论基础,建立台风灾害评估模型,并基于ArcGIS平台开发台风灾害实时评估系统,开展台风灾害的灾前、灾中、灾后评估。主要取得进展如下:(1)将长时间序列的数值天气预报数据,低精度、高时空分辨率的多普勒天气雷达数据,高精度、低时空分辨率的自动气象站数据通过数据处理,统一格式应用到台风灾害的实时评估中,为台风灾害的灾前预评估、灾中评估、灾后评估提供及时有效的信息源。(2)在研究台风灾害评估与可拓理论的蕴涵关系的基础上,详细阐述了台风灾害评估的可拓性,为台风灾害可拓评估的模型建立提供理论基础。(3)台风造成的灾情是孕灾环境、致灾因子、承灾体相互作用的结果。在海南岛台风灾害历史数据库和台风致灾因子分析的基础上,从孕灾环境、承灾体、致灾因子综合方面确定了台风灾害评估的影响因子。选择了综合孕灾因子、承灾体综合经济易损性作为静态因子,降水量、降水强度、最大风速作为动态因子,作为台风灾害评估的主要评估因子,并确定了灾害等级的划分标准。(4)探索出一套以可拓模型理论为基础,从“单元格确定→影响因素选取评估结果等级的划分→评估因子域值的划分→标准物元模型→因子权重系数确定→评估模型”的台风灾害评估方法。用可拓集合的关联函数值—关联度大小描述各种特征参数与所研究对象的从属关系,从而把属于或不属于的定性描述扩展为定量描述,使台风灾害评估从定性走向定量化,评估结果空间分辨率达到1km×1km。由于定量表达灾情的损失比较困难,在可拓理论的基础上,采用灾害等级的区间数评估模型的基础上,建立了以数值天气预报、多普勒天气雷达、自动气象站为信息源的灾前、灾中、灾后评估模型,并以不同级别来表示灾害强度;在评估模型中对台风灾害评估中的静态因子选择最优点在中点的关联函数,而对动态因子,如降水强度,降水量,最大风速采用侧距函数方法;确立了评估因子的经典域、节域范围,确立灾害评估标准物元模型;对降水和风的插值方法进行改进;对多普勒天气雷达风和降水误差进行订正;同时针对可拓算法中的缺陷,对模型进行改进,解决了当待评估数据超出某一指标评价指标范围,其关联度函数就会出现无法计算的情况,使评估模型更准确地反映实际灾情损失程度。(5)建立了台风灾害动态时空数据库,以海南岛历史台风灾害数据库为基础,利用统计学方法确定了确立了模型中的各参数,将GIS技术与台风灾害评估模型相结合,利用COM技术和ArcObjects组件,开发了基于多种信息源的台风灾害实时评估系统,实现了台风灾害的实时评估,实现了台风灾害信息化与可视化,极大提高了台风灾害防灾减灾决策的效能。(6)通过建立的台风灾害实时评估系统,以2009年的影响海南岛的热带气旋“天鹅”为例,开展了应用研究,并与实际结果对比分析,验证了系统的可行性。

【Abstract】 Survive and development of human were accompanied by natural disasters. typhoons, earthquakes, floods, thunderstorms and tornadoes, snow storms, avalanches, volcanic eruptions, heat waves, mountain landslide (debris flow), tidal (tsunami) were 10 types natural disasters threating to human survive by the World Meteorological Organization. Death toll caused by the typhoon was the highest in 10 types natural disasters, especially in Asia. It was necessary for us to carry out the typhoon disaster damage assessment. For hazard assessment was a comprehensive reflection of the disaster, not only the fundamental basis for mitigation decision-making, but also social and economic development plan.Typhoon disaster assessment was focused on direct economic loss assessment usually. Direct economic loss assessment was calculated by variously movable and immovable property with the conventional method, which was lack of real-time, for the result of economic loss were usually calculated after few days, weeks or months. The assessment of economic losses was relatively difficult for lacking of basic statistical data and not a effective way to estimate the impact in time scale. By the number of disasters, disaster losses was composited by interaction of factors, some of them can be measured by the precise mathematical model, while others not be described by accurate mathematical model,whose value may change within a certain range, so it was difficult to get an accurate value only can be give a rough estimate. For typhoon was a dynamic complex system, its structure, strength, path, wind and rain at various stages will produce a series of dramatic changes in the same location at different times. We must quickly grasp at disaster losses in the process of disaster prevention and mitigation of typhoon disaster, but the traditional statistical methods was effectiveness in the typhoon disaster emergency, so we must explore new methods of disaster assessment .Therefore, the typhoon’s real-time dynamic assessment was the key to good disaster prevention and mitigation.There were some problems existing in typhoon disaster assessment, typhoon real-time evaluation system based on ArcGIS platform was developed by extension method with the numerical weather prediction, automatic weather station, doppler radar as the information sources to carry out typhoon disaster assessment. Main progresses were as follows:( 1) The long time series of numerical weather prediction data, low accuracy, high spatial and temporal resolution of radar data, high precision, low spatial and temporal resolution of the automatic weather station data were processed to unified format to the real-time assessment of typhoon disasters, which were provided timely and effective information sources for pre-disaster ,disaster in the evaluation, after-disaster assessment.( 2) Extension of the typhoon Disaster assessment was elaborated in detail based on researching the implication relationship between typhoon disaster assessment and extension theory, which provides a theoretical basis for establishment the extension model of typhoon.(3)The typhoon disaster was caused by disaster pregnant environment, disaster factors, disaster bodies. Typhoon assessment factors were identified from disaster pregnant environment, disaster factors, disaster bodies based on typhoon disaster in the history database of Hainan Island. The comprehensive pregnant factor, economic vulnerability as static factors, precipitation, rainfall intensity, maximum wind speed as a dynamic factors, which were chosen to main factors of typhoon disaster assessment and determined the division of disaster level.(4)A method of typhoon Assessment was put forward based on the extension theory, from "Selecting cell→determining the factors→rating of result→determining threshold→standard matter-element model→weight of factors→assessment model.”Correlation associated with the extension function value was used to describe affiliation between various features with the studied parameters, various features of correlation belonging or not belonging to were extended to quantitative description, so typhoon assessment begin from qualitative to quantitative assessment, the spatial resolution was 1km×1km. It was more difficult in quantitative expression the loss of the disaster. Different levels of hazard intensity were represented by the level of the interval number of disaster assessment model,which were established by numerical weather prediction, automatic weather station, doppler radar as the information sources.The mid-point correlation functions were applied in static factors, the side functions in dynamic factors, such as precipitation intensity, precipitation , the maximum wind speed; factor domain, domain-wide section were applied to establish standards element model of disaster assessment; The precipitation and wind interpolation method were improved. For extension algorithm in defect, the model was improved more accurately to reflect the actual degree of the disaster losses.(5)Dynamic spatial and temporal database was established.The various parameters were determined by statistical methods based on typhoon disaster database of Hainan Island; real-time assessment system of typhoon disaster was developed based on COM and ArcObjects technology. The system realized the visualization of information of typhoon, which was greatly improved disaster prevention and mitigation of typhoon disasters, decision-making.( 6) Typhoon disasters Assessment system was applied in typhoon 2009 Goni, by compared with the actual result, the result indicated that the system was feasible in disasters Assessment.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络