ABSTRACT
Municipalities use Heat Vulnerability Indices (HVIs) to quantify and map relative distribution of risks to human health in the event of a heatwave. These maps ostensibly allow public agencies to identify the highest-risk neighborhoods, and to concentrate emergency planning efforts and resources accordingly (e.g., to establish the locations of cooling centers). The method of constructing an HVI varies by municipality, but common inputs include demographic variables such as age and income - and to some extent, metrics such as land cover. However, taking demographic data as a proxy for heat vulnerability may provide an incomplete or inaccurate assessment of risk. A critical limitation in HVIs may be a lack of focus on housing characteristics and how they mediate indoor heat exposure. To provide an objective assessment of this limitation, we first reviewed HVIs in the literature and those published or commissioned by municipalities. We subsequently verified that most of these HVIs excluded housing factors. Next, to scope the potential consequences, we used physics-based simulations of housing prototypes (46,000 housing permutations per city) to estimate the variation in indoor heat exposure within high-vulnerability neighborhoods in Boston and Phoenix. The results show that by excluding building-level determinants of exposure, HVIs fail to capture important components of heat vulnerability. Moreover, we demonstrate how these maps currently overlook important nuances regarding the impact of building age and air conditioning functionality. Finally, we discuss the challenges of implementing housing stock characteristics in HVIs and propose methods for overcoming these challenges.
