Uncertainties in Modeling Spatially-Resolved Climate Change Health Impacts

描述(由申请人提供)：为了描述人口对气候变化的脆弱性，极有必要更好地了解和量化与健康影响预测相关的规模、分布和不确定性。我们建议通过分析2000年、2030年和2050年两种温室气体排放情景下空气污染和4公里分辨率热波暴露对健康的影响(即死亡)，来研究美国东海岸与气候变化相关的人口脆弱性的空间变化。我们提议的研究将利用美国疾病控制和预防中心(CDC)资助的一个项目(《评估美国东部累积气候相关健康风险》)目前正在生成的空气污染和天气预测数据。IPCC第五次评估报告(AR5)情景被用来生成这个独特的数据集，模拟基于增强型社区气候系统模式(CCSM4)，以及社区多尺度空气质量模式(CMAQ)和天气研究和预报模式(WRF)。这一数据集中的高分辨率使我们能够研究健康影响预测中的空间异质性，这对于确定目标位置特别重要，以促进备灾工作和有效的适应战略。除了确定与气候变化相关的健康影响增加的脆弱地理位置外，我们还将研究气候变化健康影响建模的每个分析步骤带来的不确定性的大小。特别是，我们将确定四个组成部分的相对重要性--温室气体排放情景、气象和空气质量模拟、暴露-反应特征、未来人口分布预测和年龄结构。我们将结合敏感性分析和蒙特卡罗模拟来检验这样的假设，即气候变化健康影响建模的每个分析步骤引入的不确定性大小存在显著差异。不确定性分析将帮助政府机构针对气候变化造成的健康影响制定强有力和适当的应对措施，并为未来的研究分配资源。我们拟议工作的最终产品将是1)基于美国东海岸PM2.5、臭氧和热浪三种应激源的综合健康影响的特定县死亡率预测，包括它们直到2050年的时间演变，以及脆弱位置的图形表示；2)这些估计的不确定性的空间显式表示，分摊到误差源。我们的工作将在非常精细的空间尺度上提高健康结果预测的精确度，促进弱势群体有针对性地做好准备。 与公众健康相关：为了评估人口对气候变化的脆弱性，我们建议确定由于热浪和空气污染暴露而对健康影响增加(即死亡率)的脆弱地理位置。我们还将研究健康影响建模的每个分析步骤引入的不确定性的大小。

DESCRIPTION (provided by applicant): To characterize population vulnerability to climate change, there is a tremendous need to better understand and quantify the magnitude, distribution, as well as the uncertainties associated with the health impacts projections. We propose to study the spatial variations in population vulnerability related to climate change in the east coast of U.S. through analyzing the health impacts (i.e., mortality) from air pollution and heat waves exposures at 4km resolution during the decades of 2000, 2030 and 2050 under two greenhouse gas emission scenarios. Our proposed study will take advantage of the air pollution and weather projections data that are currently being generated in a project ('Assessing the Cumulative Climate-Related Health Risks in the Eastern U.S') funded by the Centers for Disease Control and Prevention (CDC). IPCC 5th assessment reports (AR5) scenarios are used to generate this unique dataset and the simulations are based on the enhanced Community Climate System Model (CCSM4), coupled with the Community Multi-scale Air Quality model (CMAQ) and the Weather Research and Forecasting model (WRF). The fine resolution in this dataset allows us to study the spatial heterogeneity in the health impacts projections, particularly important for identifying target locations to facilitate preparedness efforts and effective adaptation strategies. In addition to identifying vulnerable geographical locations with increased health impacts related to climate change, we will also study the magnitude of uncertainties introduced by each analytical step of climate change health impacts modeling. In particular, we will determine the relative importance of four components - greenhouse gas emission scenarios, meteorological and air quality modeling, exposure-response characterization, and future population distribution projections and the age structure. We will combine sensitivity analysis and Monte Carlo simulations to test the hypothesis that there are significant variations in the magnitude of uncertainties introduced by each analytical step of climate change health impacts modeling. The uncertainty analysis will help government agencies develop robust and appropriate responses to health impacts caused by climate change, and allocate resources for future research. The end products of our proposed work will be 1) county-specific mortality projections based on combined health impacts from three stressors, PM2.5, ozone and heat waves, along the east coast of U.S. including their temporal evolution up to the decade of 2050, and a graphical representation of the vulnerable locations, 2) a spatially explicit representation of the uncertainty of these estimates, apportioned to error sources. Our work will allow for improved precision of health outcome projections on a very fine spatial scale, facilitating targeted preparedness in the vulnerable populations. PUBLIC HEALTH RELEVANCE: To assess population vulnerability to climate change, we propose to identify vulnerable geographical locations with increased health impacts (i.e., mortality) due to heat waves and air pollution exposures. We will also study the magnitude of uncertainties introduced by each analytical step of health impacts modeling.