Abstract
Western Sydney experiences some of the most severe urban heat in the Greater Sydney region, yet that exposure is not shared equally. This thesis extract examines the spatial intersection of urban heat and social vulnerability across ten Western Sydney suburbs by combining satellite-derived land surface temperature (LST) with a composite heat vulnerability index built from Australian Bureau of Statistics socio-economic data, population age structure and tree canopy cover. Mean summer daytime LST ranged from 38.4 degrees Celsius in leafy, advantaged areas to 45.6 degrees Celsius in low-canopy, disadvantaged suburbs. A strong inverse relationship was found between canopy cover and surface temperature (r = -0.90), and the most disadvantaged suburbs consistently recorded the highest vulnerability scores. The findings demonstrate a clear green-cover inequity in which heat exposure compounds existing social disadvantage, and they support targeted, equity-led adaptation consistent with the WSROC Turn Down the Heat Strategy.
Introduction
The urban heat island effect describes the tendency of built environments to retain and re-radiate heat, producing air and surface temperatures markedly higher than surrounding rural land. It arises from the thermal properties of dark, impervious materials, reduced evapotranspiration where vegetation has been removed, and anthropogenic heat release, and its energetic basis has been understood since the foundational work of Oke (1982). In a warming climate, urban overheating is intensifying, and cities are now recognised as sites where climate change is experienced most acutely at the human scale (Santamouris, 2014).
Heat is a serious and frequently underestimated hazard in the Australian context. Extreme heat is responsible for more deaths than any other natural hazard in Australia, exceeding the combined mortality of bushfires, floods and cyclones (Australian Institute of Health and Welfare, 2023). The severity was illustrated on 4 January 2020, when Penrith recorded 48.9 degrees Celsius, the highest temperature ever observed in the Sydney basin (Bureau of Meteorology, 2020). Climate projections indicate that such conditions will become more common, with the number of days above 35 degrees Celsius in Sydney’s west expected to rise substantially over coming decades (CSIRO & Bureau of Meteorology, 2020).
Western Sydney is distinctive because its inland position, low tree canopy and expanses of heat-retaining infrastructure combine to produce temperatures up to ten degrees warmer than coastal Sydney during heatwaves, when the moderating sea breeze fails to penetrate inland. Critically, this burden is unevenly distributed. Newer greenfield estates in the far west, together with established suburbs carrying concentrations of disadvantage, combine high exposure with limited capacity to adapt. The result is an environmental justice concern in which the residents least able to afford air conditioning, or to live in a well-vegetated street, are those most exposed to dangerous heat.
This study aims to map and explain the intersection of urban heat and social vulnerability across Western Sydney to inform equitable adaptation. Three research questions guide the analysis. First, how does satellite-derived summer land surface temperature vary across Western Sydney suburbs? Second, how is a composite heat vulnerability index, integrating exposure, sensitivity and adaptive capacity, distributed across those suburbs? Third, to what extent do tree canopy cover and socio-economic disadvantage account for the spatial pattern of vulnerability?
Literature Review
Heat vulnerability is widely operationalised through a three-domain framework of exposure, sensitivity and adaptive capacity. Reid et al. (2009) advanced an influential mapping approach that combined heat exposure with population characteristics such as age, social isolation and household resources to identify communities most at risk. This structure has since underpinned heat vulnerability indices internationally and provides the conceptual basis for the present analysis.
Australian research has applied comparable methods to domestic conditions. Loughnan et al. (2012) mapped heat health risk across Melbourne, showing that morbidity and mortality during heat events were concentrated in areas combining older populations, social disadvantage and limited green space. Parallel work on urban microclimate establishes that vegetation is among the most effective moderators of the built thermal environment, lowering surface and air temperature through shading and evapotranspiration, and that its strategic deployment is a central lever for cooling Australian cities (Coutts et al., 2016; Santamouris, 2014). The residential dimension, including building materials and household resilience, has likewise been examined in the Australian setting (Hatvani-Kovacs et al., 2018).
Within Greater Sydney, the western region has attracted specific attention. Detailed thermal mapping has confirmed pronounced spatial variation in surface temperature across Western Sydney and linked the hottest zones to sparse canopy and dense, dark-roofed development (Pfautsch et al., 2020). At the policy level, the Western Sydney Regional Organisation of Councils produced the first regional heat strategy in Australia, framing extreme heat as a public health, infrastructure and equity risk requiring coordinated response (Western Sydney Regional Organisation of Councils, 2021). Despite this progress, relatively few studies have explicitly overlaid satellite temperature data with a socio-economic vulnerability index at suburb resolution for Western Sydney. Addressing that gap is the contribution of this thesis.
Methodology
The study area comprised ten suburbs and local government areas spanning the socio-economic and environmental gradient of Western Sydney, from disadvantaged, low-canopy localities such as Fairfield and Mount Druitt to advantaged, well-vegetated areas such as Baulkham Hills. Sites were selected to capture contrast in exposure and social composition rather than to provide exhaustive coverage of the region.
Exposure was measured using land surface temperature derived from the thermal infrared band of Landsat imagery. Cloud-free summer daytime scenes were processed to a mean LST for each suburb, providing a spatially continuous measure of surface heat that integrates the influence of materials and vegetation. Three indicators then represented sensitivity and adaptive capacity: the Index of Relative Socio-economic Disadvantage decile from the Australian Bureau of Statistics Socio-Economic Indexes for Areas, where decile 1 denotes the most disadvantaged areas (Australian Bureau of Statistics, 2021); the proportion of residents aged 65 years and over, as a marker of physiological sensitivity; and tree canopy cover, as a proxy for local adaptive capacity and microclimatic buffering.
These layers were combined in a geographic information system, joining the temperature surface to suburb boundaries and the socio-economic and canopy data. A composite heat vulnerability index (HVI) was then constructed following the exposure, sensitivity and adaptive-capacity logic of Reid et al. (2009). Each indicator was rescaled to a common 0 to 1 range using min-max normalisation across the ten study areas, and the domains were combined with weights of 0.40 for exposure, 0.35 for sensitivity and 0.25 for canopy deficit to give a final index on a 0 to 100 scale. Figure 1 summarises the workflow.
Results
Mean summer daytime land surface temperature varied by more than seven degrees, from 38.4 degrees Celsius at Baulkham Hills to 45.6 degrees Celsius at Mount Druitt (Table 1). Temperature tracked closely with tree canopy: the coolest suburbs were the most heavily vegetated, the hottest the sparsest. Across the ten areas, canopy and surface temperature were strongly and inversely correlated (r = -0.90), confirming vegetation as a dominant control on local surface heat. The composite index ranged from 17.5 to 83.0 and rose consistently as disadvantage deepened, so that the suburbs with the lowest socio-economic status recorded the highest vulnerability.
Table 1: Land surface temperature, canopy, socio-economic and demographic indicators, and the composite heat vulnerability index for ten Western Sydney suburbs, ordered from most to least vulnerable.
| Suburb / area | Mean summer daytime LST (°C) | Tree canopy cover (%) | SEIFA IRSD decile (1 = most disadvantaged) | Population aged 65+ (%) | Heat vulnerability index (0-100) |
|---|---|---|---|---|---|
| Fairfield | 44.2 | 9.8 | 1 | 12.4 | 83.0 |
| Mount Druitt | 45.6 | 8.4 | 1 | 9.6 | 82.5 |
| Penrith | 44.9 | 13.8 | 4 | 13.2 | 78.9 |
| Campbelltown | 43.8 | 14.1 | 3 | 12.9 | 73.6 |
| Liverpool | 43.5 | 11.2 | 2 | 11.8 | 73.3 |
| Auburn | 43.1 | 10.6 | 2 | 9.9 | 65.1 |
| Blacktown | 42.4 | 12.5 | 3 | 10.7 | 59.8 |
| Parramatta | 41.6 | 15.3 | 5 | 11.2 | 49.7 |
| Camden | 40.2 | 22.4 | 7 | 10.4 | 27.2 |
| Baulkham Hills | 38.4 | 31.7 | 9 | 14.6 | 17.5 |
Grouping the suburbs into vulnerability tiers clarifies the pattern (Table 2). The three severe-tier suburbs averaged 44.9 degrees Celsius and only 10.7 per cent canopy at a mean disadvantage decile of 2.0, whereas the two lower-tier suburbs averaged 39.3 degrees Celsius, 27.1 per cent canopy and a decile of 8.0. Surface temperature falls, canopy rises and disadvantage eases in step down the tiers. Penrith is a notable exception: despite a more favourable canopy and socio-economic profile, its extreme exposure and comparatively old population lift it into the severe tier.
Table 2: Heat vulnerability hotspot tiers, showing the number of suburbs and mean indicator values in each band.
| Vulnerability tier | HVI band | Suburbs (n) | Mean LST (°C) | Mean canopy (%) | Mean SEIFA decile | Mean HVI |
|---|---|---|---|---|---|---|
| Severe | 75 and above | 3 | 44.9 | 10.7 | 2.0 | 81.5 |
| High | 60 to 74.9 | 3 | 43.5 | 12.0 | 2.3 | 70.7 |
| Moderate | 45 to 59.9 | 2 | 42.0 | 13.9 | 4.0 | 54.8 |
| Lower | below 45 | 2 | 39.3 | 27.1 | 8.0 | 22.4 |
The construction of the index can be illustrated for Fairfield, the most vulnerable suburb. Each indicator is first normalised to the 0 to 1 range using the observed minimum and maximum across the ten suburbs, and the domains combined using the specified weights:
- Exposure: E = (44.2 – 38.4) / (45.6 – 38.4) = 5.8 / 7.2 = 0.806
- Sensitivity, from normalised disadvantage and age: disadvantage = (9 – 1) / (9 – 1) = 1.000; age = (12.4 – 9.6) / (14.6 – 9.6) = 2.8 / 5.0 = 0.560; S = (1.000 + 0.560) / 2 = 0.780
- Canopy deficit: A = (31.7 – 9.8) / (31.7 – 8.4) = 21.9 / 23.3 = 0.940
- Index: HVI = 100 x (0.40 x 0.806 + 0.35 x 0.780 + 0.25 x 0.940) = 100 x (0.322 + 0.273 + 0.235) = 83.0
The score of 83.0 reflects the compounding of near-maximum exposure, the deepest disadvantage in the sample and among the lowest canopy cover. The calculation also shows why age alone is a weak predictor. Baulkham Hills has the largest elderly population in the study area, yet its low exposure, high canopy and socio-economic advantage return the lowest index overall, confirming that vulnerability arises from the interaction of exposure and social conditions rather than demography in isolation.
Discussion
The central finding is one of green-cover inequity. The suburbs bearing the greatest heat exposure are systematically those with the least tree canopy and the deepest disadvantage. Because canopy so strongly governs surface temperature, the uneven distribution of vegetation across Western Sydney effectively allocates thermal risk along socio-economic lines. This pattern is consistent with Australian evidence that heat health risk concentrates in disadvantaged, low-vegetation communities (Loughnan et al., 2012) and with thermal mapping of the western region (Pfautsch et al., 2020). It carries an environmental justice dimension, since residents in the severe-tier suburbs are also more likely to rent, to face housing stress and to have limited means to run air conditioning through prolonged heat.
The adaptive-capacity gradient compounds the exposure gradient. In suburbs such as Fairfield and Auburn, residents from culturally and linguistically diverse backgrounds may face language barriers to heat health messaging, while lower incomes constrain the uptake of cooling and the retrofitting of dwellings. Where the built fabric consists of dark roofs, small lots and minimal setbacks, the scope for private cooling is further reduced. Heat therefore acts as a threat multiplier, deepening disadvantage that already exists.
The results point to several planning levers. The most direct is expanding urban tree canopy, identified here as the single strongest moderator of surface temperature and aligned with the 40 per cent canopy aspiration and Sydney Green Grid in metropolitan planning (Greater Sydney Commission, 2018). Equity should govern where new canopy is directed, prioritising the severe-tier suburbs that currently have the least. Complementary measures include cool and reflective roofing and paving to lower the thermal load of the built surface (Santamouris, 2014), water sensitive urban design to sustain evapotranspirative cooling (Coutts et al., 2016), and the embedding of heat in development controls consistent with the New South Wales Greener Places guidance (NSW Department of Planning and Environment, 2022). Coordinated delivery through the Western Sydney Regional Organisation of Councils Turn Down the Heat Strategy can target interventions where the index is highest (Western Sydney Regional Organisation of Councils, 2021).
Limitations
Several limitations qualify these findings. Land surface temperature is not equivalent to the air temperature experienced by residents, and although an effective comparative measure of exposure, it captures a single set of satellite overpass conditions rather than the elevated night-time temperatures that most strongly affect health. The index relies on a defined set of indicators and a chosen weighting scheme, and alternative weights would shift individual scores, so it is best read as a relative ranking rather than an absolute measure of risk. Suburb-level aggregation may also mask variation within suburbs. Finally, the ten suburbs were selected to illustrate the socio-environmental gradient rather than as a random sample, so the reported correlations should be read as indicative.
Conclusion
This study overlaid satellite-derived land surface temperature with a socio-economic vulnerability index across ten Western Sydney suburbs. Surface temperature varied by more than seven degrees and was strongly and inversely related to tree canopy, while the composite index rose in step with disadvantage, so that the hottest suburbs were also the least vegetated and the most disadvantaged. The evidence describes a clear green-cover inequity in which heat exposure compounds existing social vulnerability, concentrating thermal risk among the communities least equipped to manage it. Equitable adaptation, led by targeted canopy expansion and cooler urban materials in the highest-vulnerability suburbs, offers the most direct means of narrowing this gap. Future work should incorporate night-time temperature, extend the analysis across the metropolitan area and validate the index against health outcome data.
References
Australian Bureau of Statistics. (2021). Socio-Economic Indexes for Areas (SEIFA), Australia, 2021. Australian Bureau of Statistics.
Australian Institute of Health and Welfare. (2023). Heat and health in Australia. Australian Institute of Health and Welfare.
Bureau of Meteorology. (2020). Special Climate Statement 73: Extreme heat and fire weather in December 2019 and January 2020. Commonwealth of Australia.
Coutts, A. M., Beringer, J., & Tapper, N. J. (2016). Green infrastructure and the mitigation of urban heat in Australian cities. Progress in Physical Geography, 40(1), 105-134.
CSIRO & Bureau of Meteorology. (2020). State of the climate 2020. Commonwealth of Australia.
Greater Sydney Commission. (2018). A metropolis of three cities: The Greater Sydney region plan. NSW Government.
Hatvani-Kovacs, G., Belusko, M., Pockett, J., & Boland, J. (2018). Assessment of heat stress resilience in the Australian residential context. Building and Environment, 137, 39-52.
Loughnan, M. E., Nicholls, N., & Tapper, N. J. (2012). Mapping heat health risks in urban areas. International Journal of Health Geographics, 11, Article 49.
NSW Department of Planning and Environment. (2022). Greener places design guide. NSW Government.
Oke, T. R. (1982). The energetic basis of the urban heat island. Quarterly Journal of the Royal Meteorological Society, 108(455), 1-24.
Pfautsch, S., Wujeska-Klause, A., & Walters, J. R. (2020). Measuring and mapping urban heat in Western Sydney. Western Sydney University.
Reid, C. E., O’Neill, M. S., Gronlund, C. J., Brines, S. J., Brown, D. G., Diez-Roux, A. V., & Schwartz, J. (2009). Mapping community determinants of heat vulnerability. Environmental Health Perspectives, 117(11), 1730-1736.
Santamouris, M. (2014). Cooling the cities: A review of reflective and green roof mitigation technologies to fight heat island and improve comfort in urban environments. Solar Energy, 103, 682-703.
Western Sydney Regional Organisation of Councils. (2021). Turn down the heat strategy and action plan. Western Sydney Regional Organisation of Councils.