Heat-related illness resulting from extreme heat events is a significant health risk for residents of urban areas. One in five deaths caused by natural hazards in cities results from heat-related illness (Jesdale, Morello-Frosch and Cushing, 2013) and national-scale research indicates the most heat-vulnerable areas are city centres (Reid et al., 2009). Extreme heat events are projected to be more severe, last longer, and occur more frequently in the next twenty years in North America (Martin et al., 2012), and the health impacts of these events, including heat-related illness and death, require municipalities to mitigate extreme heat (Reid et al., 2009).
Within urban populations heat-related illness is disproportionately distributed, and vulnerability to heat is highest among several key subpopulations. Infants and those over 65, people with low income or limited education, people with pre-existing health conditions, racial minorities, and people with limited access to cooling infrastructure (e.g. air conditioning, pools) are most vulnerable to sustained high urban temperatures (Ho, Knudby and Huang, 2015). The residences of these heat-vulnerable people correlate spatially with urban environments high in heat-retaining impervious surfaces and limited tree-cover (Harlan et al., 2006). Heat-vulnerable residents have limited capacity to manage their own temperature during extreme heat events and would benefit from increased neighbourhood-level heat mitigation through improved tree-cover (Reid et al., 2009).
Researchers identify similar factors relating socioeconomic characteristics to heat vulnerability across many North American cities (Stewart et al., 2017; Ho, Knudby and Huang, 2015; Jesdale et al., 2013; Reid et al., 2009). National-scale heat-vulnerability research demonstrates general socioeconomic vulnerability trends, but Reid et al. (2009) suggest that socioeconomic variables may interact differently over the larger scale of urban centres. As such, in order to produce data-driven municipal heat mitigation strategies, spatial analysis of heat vulnerability requires municipal-level data. This scale of research was successfully conducted by Aminipouri, Knudby and Ho, (2016) and O’Neill, Zanobetti and Schwartz (2005); however, no research has been done to relate social vulnerability to neighbourhood tree cover in the study area, Montreal, Quebec. Price et al. (2013) analyzed the City of Montreal’s heat response plan during a 2010 heat wave in the city and found 106 possible heat-related deaths. Following this event, the Montreal heat response plan needed to be updated to facilitate communication to vulnerable residents (Price et al., 2013). Pham et al. (2017) posited that low-income neighbourhoods in Montreal could benefit from more tree planting.
Geographic Information Systems (GIS) are critical for analysing the spatial distribution of socioeconomic and biophysical components of heat vulnerability. Using GIS and remote sensing allows for more effective spatial analysis across large areas, with minimal cost and time (Chowdary et al., 2017). GIS tools provide an effective method for dealing with large datasets and enable the combination of heterogeneous data types. This is important to the analysis as this study deals with both socioeconomic and environmental data. LiDAR data is used to assess tree cover throughout entire cities, eliminating the need for labour-intensive and costly tree-data collection (Alonzo et al., 2014; Zang and Lui, 2012; Ejers et al., 2016; Hofman et al., 2014).
GIS becomes particularly important for municipal planning efforts where there is a limited budget as it allows for visualization of projects and value-added decision making (Coutinho-Rodrigues et al., 2011; Webster, 1993). A municipality with a pre-determined budget for increasing tree cover can use the methods outlined here to identify areas of greatest vulnerability and target these areas for tree-planting for the most effective use of the budget.