Could urban greening mitigate suburban thermal inequity?: the role of residents’ dispositions and household practices

Over the past decade research on urban thermal inequity has grown, with a focus on denser built environments. In this letter we examine thermal inequity associated with climate change impacts and changes to urban form in a comparatively socio-economically disadvantaged Australian suburb. Local urban densification policies designed to counteract sprawl have reduced block sizes, increased height limits, and diminished urban tree canopy cover (UTC). Little attention has been given to the combined effects of lower UTC and increased heat on disadvantaged residents. Such impacts include rising energy expenditure to maintain thermal comfort (i.e. cooling dwellings). We used a survey of residents (n = 230) to determine their perceptions of climate change impacts; household energy costs; household thermal comfort practices; and dispositions towards using green infrastructure to combat heat. Results suggest that while comparatively disadvantaged residents spend more on energy as a proportion of their income, they appear to have reduced capacity to adapt to climate change at the household scale. We found most residents favoured more urban greening and supported tree planting in local parks and streets. Findings have implications for policy responses aimed at achieving urban climate justice.


Introduction
Global patterns of urbanisation have concentrated people in cities (Seto et al 2010, Roberts 2011 at a time of escalating climate change, with heatwaves increasingly impacting many cities across the planet (Mcgeehin and Mirabelli 2001, Tong et al 2014, Vardoulakis et al 2014. Cities magnify heat through built form and reduced vegetation cover. This urban heat island (UHI) effect results in substantially higher temperatures in the urban core than in suburbs and hinterlands (Harlan et al 2007, Mccarthy et al 2010, Maller and Strengers 2011, Xiang et al 2014. As a result of uneven social geographies, urban heating disproportionately impacts lowerincome and ethno-racially marginalised populations -a phenomenon termed 'thermal inequity' Chakraborty 2014, Mitchell andChakraborty 2015). Such populations can become spatially concentrated in hotter urban environments (Harlan et al 2007), and may not be able to afford to cool their homes due to lower-incomes, an energy security concern (Byrne and Portanger 2014). Thermal inequity has emerged as an important climate justice issue in recent research.
Climate justice refers to efforts to overcome the inequitable distribution of climate change burdens and benefits among populations, and also to actions intended to remedy unfair responsibilities for mitigation and adaptation at the national scale (Adger 2001, Duus-Otterstrom andJagers 2012). Proponents of climate justice contend that steps must be taken to limit environmental inequities stemming from climate change, especially in cities (Stone et al 2012, Battaglia et al 2014. For example, actions to mitigate thermal inequity include urban greening (hereafter green infrastructure) designed to reduce the disproportionate impact of urban heat on marginalised and vulnerable populations (Wolch et al 2014, Byrne et al 2015. Researchers have found that green infrastructure, such as street trees and parkland, can significantly reduce direct and ambient temperatures (Lafortezza et al 2009). Although there has been some attention to modelling thermal inequity at the metropolitan scale, using remote sensing for example, less attention has been given to understanding how urban residents experience thermal inequity at the local scale, especially energy costs associated with cooling (Byrne and Portanger 2014). Little policy-oriented research has considered how green infrastructure might mitigate thermal inequity (Gill et al 2007, Hamin and Gurran 2009, Gaffin et al 2012, Norton et al 2015. Fewer studies have assessed how residents' environmental values might shape their disposition towards urban greening for climate justice (Kirkpatrick et al 2012). And less is known about suburban contexts.
Gold Coast City, like other Australian cities, is presently taking steps to reduce urban sprawl by reducing lot sizes, increasing building heights and focusing development within existing urban footprints. For many Australian cities, UTC is highest on private, not public land (Shanahan et al 2014). Urban consolidation policies often reduce UTC (Brunner and Cozens 2013), potentially leading to inequities in access to ecosystem functions, services and benefits. In the context of increasing urban heat-associated with changes to both built form and local climate-it is important to better understand the processes that create thermal inequity and how best to gauge the efficacy of potential local-scale policy interventions, such as green infrastructure (Jenerette et al 2011).
This article reports the results of research examining suburban residents' awareness of heat impacts associated with climate change and their receptiveness to urban greening as an adaptive response. By examining residents' self-reported awareness of climate change, level of concern about associated risks (e.g. heat), and perceived capacity to adapt, the study expands knowledge of thermal inequity at the household scale. Three objectives informed the research: (i) to determine if a comparatively socio-economically disadvantaged population in a suburban environment with low urban tree canopy cover may be inequitably exposed to heat; (ii) to determine if this population is disproportionately impacted by energy costs and thus has less capacity to adapt (e.g. via air-conditioning); and (iii) to determine if this population is favourably disposed towards using green infrastructure to mitigate heat exposure. The study extends existing research by focusing on the household scale, a suburban locale, and on residents' perceptions.

Data and methods
2.1. The case study area: Upper Coomera, Queensland, Australia The presence of thermal inequity rests on two conditions: (i) greater exposure of a population to climate change related temperature increases and/or to UHI effects and (ii) the lower-socio-economic and/or ethno-racial minority status of that population. We selected a single case study based on an assessment conducted by the City of Gold Coast Council that the suburb exhibited socio-economically marginalised residents vulnerable to heat stress due to social and physical characteristics (Gold Coast City Council 2011a). This locality is within a rapidly expanding city. Planning policies directed at curtailing sprawl are increasing heat-island effects. As we discuss below, land clearing prior to development removed tree canopy cover. The built form of the neighbourhood-including roof colour, building materials, yard sizes, and building density-now traps heat and without greening, will continue to do so in the future. The unit of analysis is a suburb, as defined by the local municipality, which is a sub-area of the larger statistical local area of the same name.
Upper Coomera is situated in the northern corridor of Gold Coast City, South East Queensland (SEQ). SEQ accommodates 70% of Queensland's total population and is one of Australia's fastest growing metropolitan areas. Gold Coast City was the largest contributor to population growth in SEQ between 1991 and 2013 (Queensland Treasury 2015 p 5). Council is presently investigating feasibility of using green infrastructure to combat heat island effects as part of its draft Urban Greenspace 2030 planning strategy. The study area is bounded by four major roads. Reserve Road to the south intersects with Old Coach Road along the western boundary (see figure 1). To the north is Days Road, which intersects with the M1 Pacific Motorway to the east. The topography of the study area is gently undulating, with the land becoming increasingly steep towards the western boundary (Gold Coast City Council 2011b, p 1). The neighbourhood is bisected by a vegetation corridor, adjacent to Yaun Creek, important to state and locally significant species such as Koalas (Department of Natural Resources and Mines 2015).

Climate and socio-demographic characteristics of Upper Coomera
The climate of Gold Coast City is warm, humid subtropical, with average summertime temperatures of 28°C (82 F) and winter averages of 21°C (70 F). There are localised temperature variations with suburbs further inland (including Upper Coomera) experiencing hotter and more humid temperatures. The maximum recorded temperature is above 40°C, and relative humidity above 85%-conditions conducive to heat stress (Gaffen and Ross 1998). Granger and Hayne (2001) note that South-East Queensland, including Gold Coast City, is particularly vulnerable to heatwaves. The Commonwealth Scientific and Industrial Research Organisation (CSIRO 2016) climate analogues website shows that Gold Coast City will continue to experience heatwaves as climate change intensifies in the coming decades.
The current population of Upper Coomera is 21 136 people (Australian Bureau of Statistics 2013). The suburb's socio-demographic profile exhibits several markers of social vulnerability generally, and vulnerability to heat specifically. Almost three quarters (71%) of residents identified as a couple or single parent household with children (Australian Bureau of Statistics 2013). Children (0-14 years old) comprise almost a third (29%) of the suburb's population (Australian Bureau of Statistics 2013). A small proportion of elderly residents (5%) also reside in the suburb.
Younger children and older people are especially vulnerable to heat (Maller and Strengers 2011). Children adjust to changes in environmental heat more slowly than adults. Elderly people (65 years or more) can also be prone to heat stress, due to pre-existing medical conditions and prescription medicines that may impair temperature regulation (Centres for Disease Control and Prevention 2015). Socio-economic disadvantage can reduce household capacity to afford electricity for cooling, and may reduce access to energy efficient appliances, heightening thermal inequity (Moore et al 2016). Occupation may also exacerbate thermal inequity. For example, trade workers who spend extended periods of time outdoors can be exposed to dangerous levels of heat (Ohs Reps 2015).
The Australian Bureau of Statistics (ABS) has developed the Socio-Economic Indices for Areas (SEIFA) to assess comparative disadvantage (Australian Bureau of Statistics 2014). We assessed SEIFA in the study area using a geographic information system. Pockets of concentrated disadvantage exist within Upper Coomera (figure 2), as well as broader areas of comparative disadvantage (with some pockets of comparative advantage). Compared to the state average, Upper Coomera exhibits fewer residents with 'white collar' jobs and more technical and trade workers (Australian Bureau of Statistics 2013). Based on the aforementioned criteria, the study area can reasonably be characterised as possessing many of the key sociodemographic indicators of thermal inequity.

Built environment characteristics of the study area
Absorption of solar energy on roofs is influenced by roofing colour and material and is a key factor in determining the intensity of the urban heat island effect (Watkins et al 2007, p 90). So too are reduced tree canopy cover, smaller lot sizes, built form, and higher population density. Energy consumption within individual buildings is affected by design. The energy efficiency of a dwelling can, in turn, affect the health of its residents (Younger et al 2008, p 520). We assessed the composition of roof colour in the study area using high-resolution aerial photographs. The built form of the study area is comprised of mainly detached brick-veneer houses with tile roofs (68% of the building stock), which line narrow interlinking streets. Almost a fifth (17%), of the dwelling stock is comprised of duplexes, positioned on the corners of culs-de-sac. There are also three townhouse complexes (14%) and several low-rise apartment buildings (1%), dispersed throughout the study area. A third (34%) of residential buildings in the study area have dark roofs. Based on this assessment, the study area contains built environment features (higher density dwellings with brick construction and dark roofs) that, when combined with social disadvantage, can produce thermal inequity.

Survey design
A mail-back survey was undertaken in collaboration with the City of Gold Coast Council. The survey instrument was adapted from a previous study by Byrne et al (2015), which focused on parks and longterm climate change impacts in Hangzhou, China. That study examined similar issues that we sought to assess. The modified survey instrument consisted of 43 questions, divided into four parts: (i) urban greening, (ii) views of climate change, (iii) use of neighbourhood parks, and (iv) socio-demographic measures. The instrument included measures for walkability, neighbourhood support, environmental values and a thermal comfort index. Questions measuring energy use, energy type, and energy efficiency were included. An intercept pilot survey was conducted to test instrument efficacy, and the pilot data were used to check measures. Completion time was estimated between 15 and 18 min. The research protocol was approved by the home institution's Human Research Ethics Committee (ENV/07/15/HREC). The survey design, based on the Dillman technique, employed steps to increase response rates (Dillman et al 2012). A pre-notice letter was posted to all households in the study area. Next, the survey was distributed to letterboxes together with a cover letter encouraging residents to respond. Finally, a reminder postcard was mailed two weeks after the survey to thank respondents, as well as to remind non-respondents to complete the questionnaire. Surveys were distributed to all 1921 households in the study area, comprising a mix of single-family houses, duplexes (attached dwellings) and apartments.
The equation below illustrates the calculation of the ideal sample size (Selvanathan et al 2011), where n is the sample size, z α/2 is the z-critical value (1.96) for a two-tailed test, p is the expected proportion (0.50 or 50% is used in lieu of knowledge regarding the expected proportion (this conservative approach maximises the standard deviation of the estimate of p)) and E is the error of estimation (±0.05 or ±5%). The target size for a sample of residents from Upper Coomera was 196. A total of 230 surveys were returned during the 37 day survey collection period, yielding a 12% response rate. A suitable sample size was thus achieved. Some limitations must be acknowledged. Selfselection bias is inherent in mail-back surveys. Respondents are not randomly selected and there are a greater number of non-responses (Veal 1992). The sample had a higher proportion of female respondents, an older median age, and fewer renters when compared with Australian Census data (see table 1). Relative disadvantage in the suburb may thus be higher than captured by the survey. Minor wording issues were found in the survey instrument. For one question, a number of mature aged respondents incorrectly checked 'high school student' as the highest level of educational attainment, misunderstanding the response options.

Analysis
Data were entered into Survey Monkey and then exported into SPSS (v.22) and Stata/IC (v.13) for analysis. Following Field (2009), the chi-square goodness of fit test was used to determine the representativeness of the survey data for the study area (see table 1). Ordinary least squares and probit regressions were used in the analysis. The use of probit regression appreciates the binary nature of many of the dependent variables, often taking a value of 0 or 1. Variance inflation factors were found to be below the conventional rule of thumb of 10, allaying any concerns about multicollinearity among the independent variables.

Results
Using the survey data, we examined residents' awareness of climate change impacts and perceived efficacy of various responses, energy use, thermal comfort and disposition towards green infrastructure (e.g. street trees, parks, urban greenery). Nine probit models were employed to better understand what factors or variables may underpin or explain differences in respondents': (i) concern about climate change, (ii) perceptions of effective climate change responses, (iii) energy security, and (iv) energy consumption and how these vary with respondents' socio-demographic characteristics, environmental values, and dwelling type. Marginal effects are reported throughout the paper and are interpreted in terms of a percentage change in the likelihood of reporting the dependent variable outcome for a one-unit or discrete change (for example, from 0 to 1) in the independent variable. For instance, table 2 column 1 indicates that a one-unit increase in a resident's years of age (where one unit is 10 years of age) is associated with a 7.4% reduction in the likelihood of reporting that they are concerned about climate change (all things being equal). Model 1 (see table 2) sought to establish the characteristics of respondents who were more concerned about climate change. The results show that individuals living in duplexes, households with energy efficient appliances or supplied with natural gas, and people with a greater degree of anthropocentric belief (instrumental view of nature), are more likely to be worried about climate change. In contrast, households with unrelated adults, individuals living alone, couples with no children, and those with an extra 10 years of age, are less likely to be worried about climate change. Households with dark roofs or energy efficient lighting, and those who have an additional child, are also less likely to be worried about climate change. It may the case that this result reflects that people who are more concerned about climate change are already taking action to protect themselves against expected impacts.
Model 2 (table 3) explored the characteristics of respondents linked to a respondent's inclination to insulate their dwelling as a climate change response. The model shows university students, households with insulation, or individuals with more anthropocentric values (p-value=0.112) are more likely to suggest insulating their dwelling. In comparison, individuals who have an additional $1000 in annual household income are 0.4% less likely to suggest insulating their dwelling. This is a statistically significant, albeit marginal difference, which suggests that wealthier people may either already have insulation, or can afford to do without it (perhaps because they are well-placed to bear the costs associated with running their air-conditioning). Model 3 (table 4) investigated the relationship between respondents' inclination to buy energy efficient appliances as a climate change response and those characteristics of a respondent thought to potentially have a bearing on a respondent's answer. University graduates, individuals with a pool/spa or insulation, as well as those who have a higher degree of anthropocentric belief are more likely to suggest buying energy efficient appliances. However, males (p-value=0.108), individuals living in townhouses, and those who have an additional child in the household (p-value=0.103), are less likely to suggest buying energy efficient appliances. This finding could be attributable to comparatively less wealthy residents being unable to afford energy efficient appliances, given other more salient competing demands (e.g. child-rearing, mortgage payments). It is also possible that an individual's wealth, which is often tied to housing, as distinct from income, is incompletely captured by the income variable.
Model 4 (table 5), identified factors that may dispose respondents to suggest using fans instead of air conditioners as a climate adaptive response. Results indicate that males and individuals with solar hot water (p-value=0.107) or roof ventilation are more likely to suggest using fans instead of air conditioners. Conversely, couples with no children, single parents and respondents who have an additional child in the household are less likely to suggest using fans. Households with a pool/spa, individuals with an extra 10 years of age, and those who spend an additional $100 on energy per quarter (p-value=0.110) are also less likely to suggest using fans instead of air conditioning. Awareness may play a role here, with individuals already taking adaptive responses being more likely to take further adaptive actions.
Model 5 (table 6) examined the characteristics of respondents who are receptive to having a light- coloured roof as a climate change adaptation. Individuals with solar hot water or roof ventilation (p-value=0.120) are more likely to suggest using lightcoloured roofs, whereas university students (p-value=0.102) are less likely to suggest such an intervention. The results show that respondents' implementing other energy efficiency measures are more likely to suggest having light-coloured roofs as an adaptive response to climate change, perhaps because they are pre-disposed to taking energy efficiency actions. While university students may be more aware of climate change impacts (Wachholz et al 2014), they often face financial stress (whether living with their parents or renting), and/or may have less personal investment in a property, potentially impacting their receptiveness to this intervention (Watson et al 2015).
Model 6 (table 7) examined the characteristics of respondents associated with an inclination to use photo voltaic (PV) solar panels as an adaptive response. The model shows that individuals with a pool/spa, roof ventilation and solar hot water (p-value=0.108), as well as those occupying their home for an additional year, are more likely to have solar panels. On the other hand, individuals who spend an extra $100 on energy per quarter are less likely to have solar panels. Renters would be unlikely to install photovoltaic solar panels because they do not own the dwelling. Homeownership is typically necessary for solar panel installation due to necessary property rights for modifying a dwelling and to the ability to access finance (O'Doherty et al 2008, Parkinson et al 2009). Further, some households with a pool (and likely higher wealth) may be acting to offset operating and maintenance expenses by installing PV panels. Households who are unable to afford energy efficiency measures are likely to incur higher energy expenses, potentially limiting their disposable income available for adaptive responses such as PV solar panels.
Model 7 (table 8) investigated the characteristics of respondents associated with the use of insulation in their dwellings. The results indicate that individuals occupying their home for an additional year are 2.18% more likely to have insulation. While university graduates (p-value=0.102), renters, households with unrelated adults, individuals who have an additional child in the household, or those who spend an extra $100 on energy per quarter are less likely to have insulation. These findings broadly support those of Model 2 and Model 5 relating to a respondent's inclination to respectively insulate their dwelling and to have a light coloured roof, as a climate-adaptive response. Comparatively more disadvantaged households with lower disposable income appear to be located in dwellings that are more vulnerable to heat. Model 8 (table 9) sought to reveal the characteristics of respondents that are associated with a respondent's energy demand, specifically dependence upon grid electricity and the use of a pool or spa. Model 8 addresses the issue of energy security by focusing on respondents who only use grid electricity. Results generally support the findings from Model 7. The model shows that males, individuals who live in duplexes, and those who spend an extra $100 on energy per quarter are more likely to only use grid electricity.
However, individuals with a pool or spa are 37% less likely to only use grid electricity. Because households with higher energy costs seem more likely to be restricted to grid electricity, this raises concerns about energy security problems facing renters and lower-income households. They may be especially vulnerable to rising electricity costs-particularly if they reside in a dwelling with low thermal efficiency-and thus may be unable to afford to cool their dwellings (Moore et al 2016).
The final regression, Model 9 (table 10) assessed which variables may be associated with a respondent's use of a pool or spa. Results show that high school graduates, individuals with an extra 10 years of age, and those occupying their home for an additional year (1%), are more likely to have a pool or spa. Likewise, individuals living in townhouses, households with PV solar panels or dark roofs are also more likely. In contrast, renters, individuals who live in duplexes, and households with solar hot water are less likely to have a pool or spa. Some owner-occupiers and renters of low-rise apartments may have access to a pool that is on common property. Older respondents appear more likely to be homeowners, and thus to have a pool. The literature suggests that education may be strongly related to awareness of climate change (Lee et al 2015), and those respondents with only a high-school education may not have made dwelling-choices with energy efficiency and/or climate change in mind.
It is important to note here the disposition of respondents towards urban greening as a potential climate adaptation response (table 11). We found that almost two-thirds of respondents either agreed (36.2%) or strongly agreed (27.7%) with the statement that more tree planting should occur in local parks and streets, and over half either disagreed (37.35) or strongly disagreed (23.2%) with the statement that there is sufficient shade on local streets.
We also examined respondents' perceptions of the benefits and costs of trees (table 12). Many respondents reported shade (90%) as a benefit and over half (52.5%) reported temperature reduction as a perceived benefit. Most respondents (71.2%) perceived maintenance costs as a perceived disadvantage of trees. As noted above, a potential policy response to urban heat is to deploy various forms of green infrastructure to cool direct and ambient temperatures in built environments. There appears to be strong support among respondents for urban greening.

Discussion and concluding comments
This research sought to determine whether thermal inequity might exist in an Australian suburb, extending environmental justice research from North America Chakraborty 2014, Mitchell andChakraborty 2015) based primarily upon higher density and inner-city locales (Jesdale et al 2013). We tested statistical associations between indicators of social disadvantage and measures of climate change awareness, concern and perceived efficacy in adapting to impacts, as well as residents' energy expenditure, perceived thermal comfort, and disposition towards use of green infrastructure as a policy intervention to lessen heat in built environments.
Respondents were very aware of climate change; many expressed concern about anticipated impacts. As temperatures increase due to climate change, more suburban households in the case study area will likely experience thermal discomfort (Holmes and Hacker 2007). An expected response would be the use of air-conditioning for cooling and thermal comfort. Indeed, there is already a 93% uptake of air-conditioning in the study area. But this response could be maladaptive for disadvantaged households. Increasing electricity prices (associated with upgrading distribution networks for peak demand and climate resilience) will likely widen an existing gap between those who can afford to run air-conditioning and those who cannot (Nierop 2014, Powells et al 2014. We expected to find a relationship between income, energy security, family composition, and home ownership. Poorer people are often renters rather than owners (and may be more likely to live in low-rise apartments). For those with children, raising a child is financially demanding and can be associated with comparative financial disadvantage, especially for single parents (Cutter 2006). Poor households are more likely to be energy insecure (Byrne and Portanger 2014). Older and wealthier residents (often homeowners) may choose to use air-conditioning for perceived thermal comfort benefits, despite operating costs. Those respondents with children may regard their child's health and wellbeing as more important than energy costs, and/or be concerned with getting a good night's sleep and/or preserving household routines. As such, these respondents may be less inclined to suggest using fans instead of air-conditioners, as has been found in recent Australian research (Nicholls and Strengers 2015).
Although we found that renters have lower annual incomes, there was no statistically significant difference for residents in a 'townhouse/low rise apartment'. Nor did we find a statistically significant relationship between dwelling type and household attitudes towards energy efficiency. These results may be due to renters having little control over the appliances installed in their dwellings; it is landlords who make that decision. Landlords may act to limit financial outlays and to maximise their rental returns, seeing limited value in installing high-end, energy efficient appliances that could be damaged by tenants. Operating costs are not their concern because they are passed onto tenants (who pay for electricity). So renters may have little experience of the benefits provided by energy efficient appliances, thus explaining the finding that they do not appear to regard such appliances as an efficacious climate change response. Our results suggest that there are already emergent social vulnerabilities to heat in Upper Coomera that will be potentially worsened by climate change. The suburb exhibits financially stressed renters, lowerincome earners, trades-workers and people living in higher-density dwellings with dark roofs and no insulation. Tree canopy cover is low. This combination of social vulnerability and built form that traps heat can lead to heat stress. Heat-stress can have pernicious consequences, including increased morbidity and mortality (Maller and Strengers 2011). Studies have found that violence and aggression tend to increase during heatwaves (e.g. Smoyer-Tomic et al 2003). Our findings suggest the need to reduce energy expenditure for vulnerable residents, to maintain or improve levels of neighbourhood conviviality, and to help prevent avoidable illness and death. Using green infrastructure would be a logical policy response, although there may be challenges (Stone et al 2012, Battaglia et al 2014. Green infrastructure offers a potential remedy to thermal inequality. International studies show that tree canopy cover can be lower in comparatively disadvantaged neighbourhoods, like the one we assessed Residents of such neighbourhoods may have less access to the ecosystem services, functions and benefits of trees (Schwarz et al 2015), although some researchers have found cases where tree canopy cover is higher in poorer neighbourhoods. Yet residents in disadvantaged neighbourhoods may be concerned about tree maintenance expenses (Heynen et al 2006). We found that parts of the study area are moderately disadvantaged, with higher levels of disadvantage in some pockets. Most survey respondents identified maintenance costs as their primary concern with trees, corroborating findings from the literature. But we also found most residents were in favour of more urban greening. The majority recognised that trees provide shade, although fewer linked this with temperature reduction and thermal comfort.
Using green infrastructure as a policy response could improve the thermal comfort of residents and assist in mitigating thermal inequity. For example, studies have found that the provision of parks in urban areas can reduce ambient temperatures and increase residents' thermal comfort (Jenerette et al 2011, Gaffin et al 2012, Norton et al 2015. Recent research from cities with subtropical and warm temperate climates suggests that parks can provide effective relief from heat, especially those with good tree canopy cover (Feyisa et al 2014, Byrne et al 2015). Our study found that survey respondents identified limited tree canopy cover as a problem with their local parks. Recent research has also demonstrated that urban greening can reduce temperatures, reduce wind speed, increase property values, lessen stress and anxiety, foster walking and cycling, mitigate flooding and calm traffic (Byrne et al 2015). However, less than half of survey respondents recognised these benefits. Findings from our research point to the need for planners and tree managers to work with residents in Upper Coomera to help them appreciate the manifold advantages of urban greening. Targeted awareness raising campaigns and better citizen involvement in greening activities could yield positive dividends. The results also point to the need for comparative research, to ascertain whether our results are local particularities, or if disadvantaged residents in other cities might share similar perspectives and experiences. Our results suggest that thermal inequity is place-specific and context-dependent, manifesting differently based on built environment and socio-demographic characteristics. At the heart of thermal inequity and climate injustice is the disproportionate exposure of vulnerable communities that are least responsible for climate change and most unable to mitigate, or adapt to, its effects. A better comparative understanding of the drivers of thermal inequity at the local scale, the  efficacy of potential policy remedies in different places, and whether there are common barriers to urban greening is essential, if we are to develop climate justice in cities.