Policy-relevant indicators for mapping the vulnerability of urban populations to extreme heat events: A case study of Philadelphia

Abstract

The recent global increase in extreme heat events linked to climate change is projected to continue. The additive effect of urban heat islands from impervious surfaces and urban heat emissions (e.g., from transportation and building cooling) exacerbates extreme heat events in urban areas, exposing dense populations to extreme heat with implications for human health. Ground- and satellite-based data on urban and suburban temperatures and vegetation over a historical period can help identify temporal and geospatial trends in heat exposure. A set of indicators has been developed to map the exposure, social sensitivity, and vulnerability of urban populations to heat wave health impacts. Guided by an Advisory Group of local planners in the pilot city of Philadelphia, localized trends of increasing urban extreme heat events using MODIS Land Surface Temperature (LST) data, confirmed with urban and non-urban temperature monitor data were identified. For the Philadelphia study area, the number of heat-event days in the urban setting has increased from approximately 4 days in 1980 to almost 12 days in 2013, while the non-urban setting has consistently experienced 5 days of heat events per year across the time period. Warmer micro-climates with limited vegetative cooling and elevated LSTs were also identified. The exposure indicator was combined with areas of high social sensitivity (e.g., low-income and elderly) to create a vulnerability indicator, showing significant overlap between highly exposed and highly sensitive populations. As a measure of the adaptive capacity of local governments to reduce the urban heat island, evidence of targeted vegetation increases or reduced localized temperatures linked to urban greening and cooling programs were sought, though none were of a scale to be identified by the 1 km satellite data utilized. The indicators have helped local decision makers to understand patterns of vulnerability, and may be used in the future to target adaptation actions and measure results (LST reduction or vegetation increase) from existing adaptation actions.