The predictability of high-risk zones for heat-related death within seven U.S. cities

The human response to high heat and humidity conditions varies from individual to individual, but over large populations there is clear evidence that extreme summertime conditions lead to elevated mortality and morbidity rates. This information has contributed to the adaptation of a number of measures aimed at protecting public health when dangerous conditions are forecast, including heat-health watch-warning systems. There is, however, a growing capacity to improve specificity in assessments of heat-related risk across both time and space, which could lead to more efficient allocation of resources associated with the activation of heat emergency management protocol. These improvements should be accompanied by an understanding of the variability in the spatiotemporal response to extreme temperatures, with particular emphasis on the predictability of geographic zones and specific time periods where the risk is greatest.

Within metropolitan areas the heat-health risk is expected to vary spatially as the urban heat island and related microclimates create different levels of exposure for residents of different communities. Similarly the underlying risk of the population varies as the demographic profile of city residents is spatially heterogeneous. A few studies have begun to explore this variability, identifying zones that might theoretically be associated with higher risk, or more recently have identified high-risk zones using observational records.

The heat-mortality response also varies across time. Although considerable debate continues regarding future climate change and population adaptation, the increase in temperatures in recent years combined with urban heat island impacts, has resulted in individuals being more frequently exposed to thermally stressful conditions in some locations. At the same time, infrastructure improvements including the use of air conditioning and building design might be decreasing the risk for some. In addition to these long-term trends, there is evidence that warm-season mortality might be influenced by antecedent wintertime conditions, and further evidence suggests that the characteristics of individual heat events including their timing, duration, and intensity, impact the number of impacted people.

Using nearly thirty years of high-resolution mortality records from seven cities spanning the climate zones of the U.S. we have found that it is generally more feasible to predict where heat-related deaths are more likely rather than when they are more likely to occur. This finding suggests increased emphasis on longer-term measures aimed at reducing the heat-health burden that might include building weatherization and educational campaigns.