Forecasting Risk in Rapidly Changing Urban Environments

2008年，城市居民人数首次超过农村居民。到2030年，城市人口将是1970年的两倍。人口的这种变化也导致了风险格局的变化，城市成为全球风险的主要来源。目前的风险评估模式不足以从迅速变化的当地风险和脆弱性的角度来描述城市环境的时空动态。我们建议开发一种定量的方法来预测快速变化的城市环境中的风险。所提出的方法的两个主要组成部分是：（a）动态的时间依赖性暴露模型和（B）时间依赖性脆弱性模型，考虑随着时间的推移，影响建筑环境的脆弱性的主要过程：结构退化，建筑物的增量扩张，不断发展的建筑实践。通过本研究开发的模型和先进技术包括随机空间和时间暴露方法，用于城市分析的新型遥感技术，统计模式识别方法和增量建筑扩建和脆弱性的概率分析，以实现随着时间的推移的风险预测。拟议的研究将实现与地区人口和城市建筑增长模式一致的风险预测暴露在极端事件中。所开发的工具将使决策者、市政府和规划者能够采取措施减少未来的风险。研究结果将被纳入当前的风险评估模型，如HAZUS，FEMA和国土安全部用于规划和管理灾害的软件，以及全球地震模型（GEM是一个全球性的非营利性地震联盟，开发地震灾害和风险评估的开源工具），这将使他们能够更好地建模和预测城市风险的复杂动态。该研究的许多组成部分将进一步成为斯坦福大学新的跨学科城市风险本科课程的一部分，由研究小组成员授课。项目团队在与代表性不足的群体合作方面有着良好的记录，并积极参与斯坦福大学的工程多样性计划，该计划为代表性不足的群体的学生提供教学，研究和指导机会。本科生将通过NSF的REU项目和斯坦福大学的研究本科生项目被纳入该项目，该项目由本科教育副教务长（VPUE）支持。

In the year 2008 urban dwellers for the first time ever outnumbered rural ones. By 2030 twice as many people will live in cities than in 1970. This shift in population has also led to a shift in the landscape of risk, with cities becoming the major source of global risk. Current risk assessment models fall short in characterizing the spatial and temporal dynamics of the urban environments in terms of rapidly changing local exposure and vulnerability. We propose to develop a quantitative approach for forecasting risk in rapidly changing urban environments. The two main components of the proposed approach are (a) dynamic time-dependent exposure models and (b) time dependent vulnerability model that consider the main processes influencing the vulnerability of the built environment over time: structural deterioration, incremental expansion of buildings, and evolving building practices. The models and advanced technologies to be developed through this research include stochastic spatial and temporal exposure methods, novel remote-sensing based techniques for urban analysis, statistical pattern recognition methods and probabilistic analysis of incremental building expansion and vulnerability to enable risk forecast over time.The proposed research will enable risk forecasts that are consistent with the population and urban building growth patterns of regions exposed to extreme events. The tools developed will enable policy-makers, municipal governments and planners to take steps towards reducing future risk. Research results will be integrated into current risk assessment models, such HAZUS, the software used by FEMA and the Department of Homeland Security for planning and managing disasters, and the Global Earthquake Model (GEM is a worldwide nonprofit earthquake consortium developing open source tools for earthquake hazard and risk assessment) that will enable them to better model and forecast the complex dynamics of urban risk. Numerous components of the research will further be part of a new Stanford interdisciplinary undergraduate class on urban risk, taught by research team members. The project team has a strong record in working with underrepresented groups and actively participates in Stanford's Engineering Diversity Program that provides teaching, research and mentoring opportunities for students from underrepresented groups. Undergraduates will be included in the project through NSF's REU program and Stanford's research to undergraduate students program supported by the Vice Provost on Undergraduate Education (VPUE).