An Environmental Justice Lens for Measuring Neighborhood Scale Vulnerability

The magnitude and location of population at risk to extreme events and climate hazards has typically been estimated by using socioeconomic indices based on census geographies However, these approaches do not capture the variability of vulnerabilities within these geographic areas, nor are they flexible enough for elucidating distributional aspects of a neighborhoods risk to current and future climate hazards. To ensure that social equity and environmental justice consequences can be fully incorporated into current and future climate impact assessments, we need to understand what influences the resilience of a specific neighborhood populations to extreme events in terms of hazard, vulnerability, and adaptive capacity. This research focuses on understanding climate change impact at neighborhood scale by accounting vulnerability and adaptive capacity of the neighborhood and techniques to delineate neighborhoods-at-risk over the landscape in terms of factors contributing to the vulnerability and adaptive capacity of neighborhoods. Our approach consists of a two main steps. In step one, we differentiate neighborhoods by clustering of building outlines derived from high resolution satellite imagery using machine learning techniques. In step two, we use populations derived from the Public Use Microdata Sample (PUMS) to disaggregate socio-economic factors like race, age, sex, poverty status, and housing using the maximum entropy dasymetric modeling method enabling us to classify neighborhoods based on vulnerabilities shared among their residents. Results from these two stages of analysis are synthesized into a typology characterizing neighborhoods in the city of Atlanta, GA by a common set of built environment and socio-economic characteristics. Once such neighborhoods-at-risk over the landscape have been identified, we can establish a scenario assisted future neighborhoods-at-risk and explore the distributional impact of projected impact of climate change in the future. Therefore, neighborhoods-at-risk can be used as an operational unit of risk over the landscape for evaluating the consequences of current and future climate hazards.