the impact of surface cover & tree canopy on air temperature ...

THE IMPACT OF SURFACE COVER & TREE CANOPY ON AIR TEMPERATURE IN WESTERN SYDNEYOCTOBER 2020

Western Sydney University2

WESTERN SYDNEY UNIVERSITY

AUTHORSDr Sebastian Pfautsch, Adam R Tjoelker

Urban Studies School of Social Sciences

Western Sydney University, Locked Bag 1797, Penrith, NSW 2751.

This research project has been funded by Campbelltown City Council, New South Wales, Australia.

With respect for Aboriginal cultural protocol and out of recognition that the campuses of Western Sydney University occupy their traditional lands, the Darug, Tharawal (also historically referred to as Dharawal), Gandangara and Wiradjuri peoples are acknowledged and thanked for permitting this work on their lands.

SUGGESTED CITATIONPfautsch, S., Tjoelker, A R. (2020) The impact of surface cover and tree canopy on air temperature in Western Sydney. Western Sydney University, 140 p.

DOI: https://doi.org/10.26183/bk6d-1466

©Western Sydney University.

www.westernsydney.edu.au

October, 2020.

Cover and other images from S. Pfautsch and istock.com

westernsydney.edu.au 3

This study represents a natural progression of the research described in “Benchmarking Heat Across Campbelltown, New South Wales”, which can be downloaded at: https://doi.org/10.26183/5d81cc131f041.

Increasing summer temperatures and urban development in Western Sydney present key challenges for contemporary planning. In the past decade, local populations of Western Sydney experienced record-breaking heat during consecutive years. All available climate projections point to a continuation of this menacing trend. Evidence-based strategies are necessary to improve the resilience of existing settlements against heat. New settlements must embrace all available means to become ‘heat smart’ from the get-go.

This research shows: » Increasing the area of hard surfaces and buildings leads to warming

» Increasing the area of open spaces and tree canopy leads to cooling

» When provided in equal proportions warming from hard surfaces exceeds cooling from open space

» The largest cooling benefits are generated by open space

» Increasing tree canopy cover has no effect on peak heatwave conditions, yet plays a significant role in cooling nighttime air temperatures

» The magnitude of attainable cooling effects in urban space is 0.8-1.3°C for mean summer air temperatures

» The capacity to lower peak heatwave air temperatures is very limited

» Open space and tree canopy cover can markedly reduce summer nighttime air temperatures

The deflating outcomes of planning and development practices during the past decade are visible in the new settlements across Western Sydney.

The vast majority of these settlements were constructed ignoring principles of heat-smart design. This must be reversed immediately to improve the resilience of local communities to cope with warming summer temperatures.

At the end of this report, a number of recommendations are provided to inform existing policies and improve planning practices for urban development. These are based on empirical data and can be used to build ‘heat-smart density’, where the ratio of open space to hard surfaces and buildings is 2:1 or greater.

PRACTITIONER’S BRIEF

BENCHMARKING HEAT ACROSS CAMPBELLTOWN, NEW SOUTH WALES

WESTERNSYDNEY.EDU.AU

CONTACT USWestern Sydney University,

Locked Bag 1797, Penrith, NSW 2751

RSCH5007 Cambelltown cover.indd 60-1 10/10/19 9:45 am



CONTENTSPRACTITIONER’S BRIEF 3

CONTENTS 4

1. INTRODUCTION 6

2. OBJECTIVE 10

3. METHODOLOGY 113.1 Study Sites 11

3.2 Geo-spatial Analyses 11

3.3 Data Analyses 16

4. RESULTS 184.1 Variability in Surface Types and Area 18

4.2 Effect of Surface Type on Summer Temperatures 20

4.3 Effect of Surface Type on January and February Temperatures 22

4.4 Effect of Surface Type on Heatwave Conditions 25

4.5 Individual Warming and Cooling Effects 26

5. RECOMMENDATIONS 305.1 Reduce Warming 30

5.2 Increase Cooling 31

5.3 The Role of Blue Space 31

5.4 Regulating Urban Temperatures with Tree Canopy 31

5.5 Low Capacity to Reduce Heatwave Temperatures 32

6. CONCLUSION 34

7. REFERENCES 35

8. DETAILS OF STUDY SITES 36

westernsydney.edu.au

SURFACE TYPES AND AIR TEMPERATURE

5



1. INTRODUCTIONSoil sealing, the process whereby an open, pervious surface that would otherwise be covered by vegetation is partly or completely capped with impervious surface materials like asphalt, concrete, or bricks, represents the most severe form of land conversion (European Commission 2012). This process results in greater absorption of solar energy, increased heat storage and slowed emission of stored energy. At a city-scale, this alteration of energy budgets leads to well documented Urban Heat Island Effects that leave cities warmer compared to surrounding rural land (Oke 1973). The process of urbanisation and associated soil sealing represents possibly the single most important anthropogenic impact on global surface climate (Zhao et al. 2014).

Urban Heat Island Effects have been documented for hundreds of cities in Asia (Hung et al. 2006, Roth et al. 2012), North

America (Lowe et al. 2006, Solecki et al. 2011, Wang et al. 2016a), Europe (Ward et al. 2016), and Australia (Santamouris 2015). Recently,

this effect was also documented for the local government area (LGA) of Campbelltown (Pfautsch and Rouillard 2019) (Figure 1).



7

FIGURE 1: Representation of the Urban Heat Island Effect (UHIE) for the local government area of Campbelltown during the nights of February 2019. The UHIE can be identified by the 4.5°C difference in air temperature between densely populated and rural areas. The warmer north and centre comprises of large areas covered by hard surfaces and buildings, whereas the eastern and southern regions have a low proportion of hard surfaces and buildings and thus remain cooler. The map was generated by interpolating measurements from 102 individual locations for 1-28 February 2019 (10-minute data from 21:00 until 05:00). The map is reproduced from Pfautsch and Rouillard (2019). Note: for reasons of restricted access, the eastern section of the (unpopulated) local government area was not included in this study.



summer heat were provided in the report “Benchmarking Heat across Campbelltown City, NSW”, available for download at: https://doi.org/10.26183/5d81cc131f041. The study demonstrated that local Urban Heat Island Effects can be captured and reproduced accurately using the data from a dense network of air temperature loggers. Figure 1 represents such a reproducion.

The analyses of the project revealed that urban communities living in the suburbs of Campbelltown City are exposed to extreme summer heat much more frequently than documented by nearby meteorological stations. Yet, the origin of the variability of temperature variation observed across the LGA was not assessed further. The present report closes this gap by analysing the influence of different surface types and their relative cover in the vicinity of the 102 measurement locations.

Evidence presented here, and recommendations derived from it will advance the understanding of how the spatial composition of urban and peri-urban landscapes impacts urban microclimate of a large tract of land in Western Sydney that is representative for the region. In light of anticipated urban development in this region, and the prospect of increasing summer heat, the present report offers practical solutions to improve the climate resilience of local communities. Armed with this knowledge, it will be possible to formulate evidence-based guidelines, policies and management strategies that increase the thermal performance of existing built-up space and inform planning and development of new suburbs.

the overpass by a satellite or aeroplane. Thus, LST cannot be used to analyse important thermo-dynamic processes in urban space like air movement and evaporative cooling. One example for these important processes is nighttime cooling and how the magnitude of cooling depends on the composition of urban fabrics. Thirdly, analyses of full temperature ranges of common urban surface materials are limited by the resolution of individual pixels from remote sensing products (Song et al. 2014, Wang et al. 2016b). The reason for this effect is that surface temperature will be averaged across each pixel, which results in the loss of information about location specific maximum and minimum LST.

Sophisticated assessments of remotely sensed LST can provide highly valuable information at relatively large spatial scales. However, understanding how the configuration of urban surface space impacts microclimates at the neighbourhood scale requires other techniques, such as analyses of local microclimates and related energy budgets (e.g., Coutts et al. 2007). Such microclimate studies will rely on measurements of near-surface air temperature and are the method of choice when assessing the direct effects of Urban Heat Island Effects (Schwarz et al. 2012). Assessments of near-surface air temperatures allow documenting microclimatic variation that local communities actually experience (e.g., Krayenhoff et al. 2020, Ziter et al. 2019). However, no method comes without limitations. It is well known that air temperature in urban environments can vary greatly and this variation is rarely captured due to the lack of a systematic network of meteorological stations. Research has shown that using data from such stations has a limited capacity to capture the apparent variation of near-surface air temperatures of urban landscapes (Morabito et al. 2015).

This issue was resolved by the joint project of Western Sydney University and Campbelltown City Council during the summer of 2018/19. The project collected detailed measurements of near-surface air temperatures at 102 locations within and beyond the boundaries of the LGA. Analyses of these measurements and recommendations for heat-smart urban planning and strategies to mitigate

Urban Heat Island Effects are predominately studied remotely. The necessary data are generated using instruments carried by orbiting satellites or fixed-wing aeroplanes and imported into geo-spatial information systems that can produce maps depicting the land surface temperature (LST) of cities and metropolitan regions. When LST is documented at high spatial resolution, assessments of the relationship between surface temperature and surface cover are possible by overlaying different layers of mapped information. Usually, these analyses report positive correlations between LST and grey infrastructure (i.e., the build-up urban space made up by roads and buildings) and demonstrate how an increasing cover of hard surfaces and buildings promotes urban warming (e.g., Estoque et al. 2017, Morabito et al. 2016).

Compared to hard surfaces, urban space covered by green infrastructure usually has a lower LST (Estoque et al. 2017) and different types of green surface cover like grass, shrubs or tall trees will have different cooling effects on LST (Duncan et al. 2019). Warming effects of hard surfaces and cooling effects from vegetation are usually linearly related to their relative cover (Morabito et al. 2016, Estoque et al. 2017), which indicates that LST can be influenced by reducing or expanding certain surface types. Indeed, a study found that by replacing 20% vegetation cover across central Manchester in the UK with asphalt would increase mean LST by nearly 5°C, while adding 5% green space in the form of mature trees would reduce LST by about 1°C (Skelhorn et al. 2014). This is a striking example of how urban surface cover influences thermal conditions in built-up space. Nevertheless, using LST to analyse urban heat has three main disadvantages.

Firstly, LST only allows an indirect measure of Urban Heat Island Effects, because it is using surface temperature data. Analysis of LST will lead to calculation of temperatures that include many locations where humans would not regularly spend much time (e.g., roofs, roads). Secondly, using remotely sensed data will provide a static representation of the thermal environment during the time of

https://doi.org/10.26183/5d81cc131f041

https://doi.org/10.26183/5d81cc131f041



9



The objective of this study was to provide novel empirical evidence demonstrating how common urban surface materials influence microclimate around the LGA of Campbelltown City.

Special emphasis was put on analysing the proportion of five surface types and tree canopy cover above these surfaces at very high resolution to develop accurate relationships between surface cover and a range of summer temperature derivatives (mean, maximum, minimum, daytime and nighttime temperatures).

While results are based on analyses using 102 locations dispersed systematically over the entire LGA of Campbelltown City (Figure 2), results generated here have a far wider relevance to urban planning and development across the region.

2. OBJECTIVE

FIGURE 2: The network of research locations across the local government area of Campbelltown. Each yellow dot indicates the position of a single temperature logger (n = 102). Note: for reasons of restricted access, the eastern section of the (unpopulated) local government area was not included in this study.



11

3. METHODOLOGYTree canopy cover as sixth category was determined separately as layer that would sit above one or more of the surface types (Figure 4).

The following procedure was used to determine the area of each surface type:

1. A 50 m circle was drawn using the location of the logger as central reference point; this object was produced using the circle function in Nearmap. Surface area of the circle covered 7,854 m2.

2. The polygon tool was used to trace the outline of the circle. This was necessary, as other polygons could only be drawn upon polygons, but not objects produced using the circle tool (internal limitation of Nearmap). Redrawing the circular area resulted in slight deviations of ±1.4% from the ideal circular area.

3. Individual polygons that delineated each plot or parcel of the five surface types were drawn manually and their area was noted. If areas under tree canopies were either covered by unvegetated or vegetated surfaces was determined by assessing the quality of ground cover vegetation in the selected map, and when necessary also in other Nearmap images representing similar summer conditions in previous years. Where surrounding ground vegetation looked sparse (e.g., in native bushland around Holsworthy), all surface type was quantified as type 2 ‘unvegetated space’.

4. The absolute area of each surface type was calculated by summing the individual area of polygons for each of the five categories. The relative area of each surface type was calculated by dividing its summed area by the area of the entire circular area, multiplied by 100.

5. Finally, polygons delineating tree canopies were drawn on top of all surface type polygons. Shrubs with a surface area <4m2 were not counted at tree canopy. Larger shrubs with tree-like characteristics were considered canopy.

3.2 GEO-SPATIAL ANALYSESThe exact position of a tree or light pole was determined in Nearmap (Barangaroo, Sydney, Australia). This subscription-only software provides access to high-resolution aerial images that cover Western Sydney, including the Campbelltown region. Regular flights are conducted that capture the area at 5.8-7.5 cm ground sample distance. The resulting maps are typically accurate to ±15 cm for distances up to 200 m. More information about the accuracy of the maps used for the analyses here can be found at: https://docs.nearmap.com/display/ND/Accuracy. The map produced by an overpass in early March 2019 was used for the analyses presented here.

At each of the 102 locations, the absolute and relative area of five different surface types as well as tree canopy cover was determined manually (Figure 3). The minimum area of individual objects and tree canopies included in these analyses was less than 2 m2. This high level of precision produced very accurate estimates for each urban surface type. Where necessary, Google Street View was used to identify surface types obscured by tree canopies. Individual objects and surface polygons were clustered into five surface types:

1. Vegetated, green open space (lawns, pastures, shrubs, playing fields, gardens, vegetated median strips, parks,etc.)

2. Non-vegetated open space (bare soil, dry ground vegetation, etc.)

3. Hard ground surfaces (roads, carparks, pavements, paths, driveways, etc.)

4. Buildings (represented by their roof tops)5. Blue space (swimming pools, waterways,

etc.)

3.1 STUDY SITESFor the 2019 study that assessed variation in near-surface air temperatures across the LGA, data was collected at 102 sites from Edmondson Park in the north to Wedderburn in the south of the Campbelltown City LGA. At each site, one microclimate data logger was installed, usually on a branch of the lower canopy of a tree (2.5-3 m above ground). In recently build suburbs where no sufficiently large tree was present, loggers were mounted to the horizontal arm of a light pole using a hydraulic lift. Details describing the data logger, its accuracy and data recording intervals were provided in Pfautsch and Rouillard (2019). Procedures for installation of the loggers can also be found in this report.

During installation of the loggers, each site was geo-tagged using satellite based global positioning information. This information was used to identify the exact position of each site (depicted in Figure 2). Sites were located in native bushland, national parks, rural and peri-urban settings, as well as industrial, commercial and residential streets in new and established suburbs. In addition, carparks, parks, playgrounds and school yards were also represented, providing the largest possible cross-section of typologies found throughout the LGA of Campbelltown.

https://docs.nearmap.com/display/ND/Accuracy

https://docs.nearmap.com/display/ND/Accuracy



FIGURE 3: Example of the detailed image analysis of surface types and tree canopy. The image shows a research location along Camden Valley Way, Edmondson Park. The orange marker indicates the geo-referenced location of the microclimate data logger, here located underneath a street tree. Polygons were drawn to measure the area of each parcel for a total of four surface types (the fifth, blue space, was not present at this particular site). Polygons of tree canopies were drawn on top of the surface polygons. Numbers indicate: 1 = vegetated open space, 2 = unvegetated open space, 3 = hard surfaces, 4 = buildings, 6 = tree canopy. Blue lines delineating polygons appear thick in the example image. This is an optical effect produced by Nearmap when zooming out of an area of interest. Zooming out was necessary to capture the full 50m radius of the assessed circular space. Polygon boarders covered less than 5 cm ground space when drawn onto the magnified map image. Image © Nearmap.



13

FIGURE 4: Representation of an urban landscape with surface type polygons only (panel A) and polygons for tree canopy overlapping those of surface types (panel B). Image © Nearmap.

It is important to mention that the condition of vegetated surfaces, especially that of turf and crop fields, changed notably during summer. This has effects on how vegetated surfaces are depicted in aerial images, where vegetated surfaces would look green in spring, when grass cover is lush with no visible patches of bare soil. Yet, depending on environmental conditions, the same vegetated surface can appear yellow and show substantial areas of bare soil at the peak or towards the end of summer.

These contrasting conditions of surface cover result in a different classification of the same surface type at different points in time (Figure 5). The associated changes in evapotranspiration, surface reflectance, absorption and re-emission of solar radiation by the area of interest is highly likely resulting in a change in near-surface air temperature, even when ambient air temperatures would be similar. The majority of the data used in the present analyses was recorded in January and February 2019. During this time the

southwestern section of the Sydney Basin received very little rain and was exposed to several heatwave events. Hence, we defined locations that were dominated by agricultural uses (e.g., around Wedderburn and Appin) and areas dominated by dry sclerophyll woodlands (e.g., sites at Holsworthy) as non-vegetated ground surfaces.



Care was taken to capture the irregular shape of canopies and not their ground-projected shadows. This was achieved by using additional images in Nearmap from different time points and sun angles. Comparison of these image series allowed accurate identification of tree canopies (Figure 6).

FIGURE 5: Example of seasonal changes of vegetated open space at a location in Menangle. Panel A: 28 December 2018. Panel B: 31 March 2019. Clearly visible is a dramatic change of ground cover vegetation from green to yellow or lush to dry. Image © Nearmap.



15

FIGURE 6: Examples of a native bushland (A) and an urban location that required assessing additional image series to identify canopy outlines in March 2019. The irregular shape of all canopies is clearly visible in these close-up representations for Sites 5 and 88.



TABLE 1: Single and composite surface types and canopy cover classes used in the assessment of near-surface air temperatures.

SURFACE TYPE

SURFACE TYPE CATEGORY DESCRIPTION

1 Single Vegetated open space

2 Unvegetated open space

3 Hard surface

4 Buildings

5 Blue space

6 Tree canopy

7 Composite Sum of vegetated open space and tree canopy (Surface types 1+6)

8 Sum of vegetated and unvegetated open space (Surface types 1+2)

9 Sum of unvegetated open space and tree canopy (Surface types 2+6)

10 Sum of vegetated and unvegetated open space and tree canopy (Surface types 1+2+6)

11 Sum of hard surfaces and buildings (Surface types 3+4)

3.3 DATA ANALYSESThe area covered by each polygon in the 102 locations was converted from square metres to the relative area of the polygon in relation to the entire sample circle. This transformation was necessary to establish equal conditions (i.e., 100% surface cover of the sample circle) for each location. The resulting data were used in regression analyses. These tests assessed the strength of the relationship between each of the surface types and canopy cover with near-surface air temperature. In addition to the five surface types and tree canopy, composite categories were generated (Table 1). Reason for the composite categories was to evaluate if the combined area of similar surface types yielded stronger or weaker relationships.

The relationships of the 11 surface types with 28 temperature derivatives were assessed using Pearson correlation analysis (Table 2). The temperature derivatives were calculated using the existing near-surface air temperature data from the 102 research sites across the LGA of Campbelltown. The derivatives covered seasonal and monthly time intervals, daytime and nighttime, the times where land surfaces undergo thermodynamic processes of heating up and cooling down, as well as five peak heatwave events (Table 2).

TABLE 2: List of temperature derivatives tested for their dependency on surface types present in the vicinity of the point of measurement.

DATA SET TEMPERATURE DERIVATIVE DATE INTERVAL COVEREDTIME INTERVAL COVERED

1 Mean summer 1 December 2018 – 28 February 2019 00:00 – 23:50

2 Max summer

3 Min summer

4 Mean daytime 10:00 – 18:00

5 Max daytime

6 Min daytime

7 Mean nighttime 21:00 – 05:00

8 Max nighttime

9 Min nighttime

10 Mean January 1-31 January 2019 00:00 – 23:50

11 Max January

12 Min January

13 Mean daytime January 10:00 – 18:00

14 Mean nighttime January 00:00 – 23:50

15 Mean February 1-28 February 2019 00:00 – 23:50

16 Max February

17 Min February

18 Mean daytime February 10:00 – 18:00

19 Mean nighttime February 00:00 – 23:50

20 Heating up January 1-31 January 2019 06:00 – 10:00

21 Cooling down January 1-31 January 2019 20:00 – 00:00

22 Heating up February 1-28 February 2019 06:00 – 10:00

23 Cooling down February 1-28 February 2019 20:00 – 00:00

24 Heatwave 1 29 December 2018 14:30

25 Heatwave 2 5 January 2019 13:10






17

A total of 308 individual models were analysed for their strength in predicting near-surface air temperatures based on the variation in relative area covered by any single or composite surface type. The strength of the resulting relationships was analysed using a density ellipse function with a 95% coverage. The strength of the resulting relationship between independent (surface types) and dependent (temperature derivatives) variables was expressed by the Pearson correlation coefficient (r, values range from -1 to +1). Statistical significance was given when the probability level (p) was below 0.05.

In addition, relationships between individual surface types and temperature derivatives were assessed using correlation analyses to estimate the level of cooling or warming that can be expected when modifying the area of a specific surface type or tree canopy. The predictive strength of these bivariate relationships was expressed by the coefficient of determination (R2). Relationships were statistically significant when p was below 0.05.



FIGURE 7: Box-whisker plot showing the relative area of the five surface types and tree canopy for all 102 analysed sites across the LGA of Campbelltown. The relative area covered by vegetated and unvegetated surfaces were summed to represent ‘pervious surfaces’, while the relative area covered by roads/carparks and buildings was combined to represent ‘impervious surfaces’. The lower and upper edge of the box represents the first and third quartile, respectively; the line inside the box represents the median, the cross shows the mean and whiskers depict minimum and maximum values. Outliers are shown as circles for each surface type where applicable.

4.1 VARIABILITY IN SURFACE TYPES AND AREAThe relative cover of a particular surface type and canopy cover varied widely among sites (Figure 7). For example, a number of sites in natural and semi-natural locations did not have any hard surfaces or buildings compared to urban sites where roads and carparks could cover more than 50% of the surface area within the measurement circle (e.g. Campbelltown, Site 50). More than 97% of the area around a measurement location in Ingleburn (Site 21) was covered by roads, carparks, walkways and buildings. Some sites at the newly build suburb of Willowdale (Sites 15 and 16) had no tree canopy at all, whereas locations at Holsworthy (Sites 25-27) and Appin (Site 85) had more than 70% canopy cover.

Among all sites, the average area covered by vegetated and unvegetated open space was 21% and 55%, respectively. At individual sites, the cover of vegetated and unvegetated open space varied from 0-100%. Areas with high

relative coverage of unvegetated open space were almost exclusively located in the south and east of the LGA, namely in Appin, Gilead, Holsworthy, Menangle and Wedderburn. Research locations in these areas were mostly located on or surrounded by paddocks, meadows, open woodlands and forests. As we examined tree canopy cover separately from open space in this study, many of the study sites in the east and south east simply had little hard surface cover. Examples depicting the range of complexity and degree of urbanisation are shown in Figure 8. Tree canopy never approached 100% cover at any site and mean canopy cover was 31%. This coverage does not represent average canopy cover across the LGA of Campbelltown City, and is certainly influenced by the fact that loggers were mounted to urban trees at nearly all sites, requiring the presence of at least one tree.

Blue space was present at 12 sites where on average it covered 3% of the entire 50 m radius measurement circle. Water covered

a relative area of nearly 15% only at Site 77 in Wedderburn and Site 95 at Spring Farm. An estimation of the total area covered by either pervious or impervious surfaces, which were the sums of vegetated plus unvegetated open space, or the sum of roads/carparks plus buildings, showed that on average 76% of the area around research locations was covered by pervious surface types, while 23% were covered by impervious surfaces. These estimates of relative surface cover provide an approximation of the investigated proportion of the Campbelltown LGA. The high presence of pervious surfaces is a reflection of the rural and natural character of the region in the south, southeast and southwest of the LGA. These analyses have also revealed that sites in newly developed suburbs like Willowdale, Bardia and Macarthur Heights boasted more than 50% area covered by hard surfaces and buildings.

4. RESULTS



19

FIGURE 8: Example for the diversity of measurement locations across the LGA of Campbelltown. (A) New urban development at Willowdale with a high area covered by buildings and zero tree canopy cover. (B) Established residential area in Bradbury with a mix of vegetated and unvegetated open spaces, roads, buildings, walkways, tree canopy and blue space. (C) Industrial estate in Ingleburn with a high proportion of buildings and hard surfaces and little vegetated green space or tree canopy cover. (D) Native bushland at Appin with tree canopy cover over sparsely vegetated open ground; for practical reasons this space was categorised as ‘unvegetated’ in the present analyses. Image © Nearmap.



When combining surface types, it becomes clear that vegetated surfaces have no additive cooling effect on urban or rural spaces across Campbelltown. The combination of vegetated open space and tree canopy did not yield any meaningful cooling benefits (Figure 9). However, when combining the unvegetated surface category with that of tree canopy, multiple cooling benefits emerged, especially for mean summer (r = -0.64, p <0.001) and nighttime temperatures (mean: r = -0.67, p <0.001; max: r = -0.72, p <0.001; min: r = -0.64, p <0.001).

The largest cooling effect on summer temperatures was found for blue space.

This category included ponds, streams and swimming pools at 12 of the 102 sites. Coefficients of determination for this surface type were as high as 0.95, indicating a near ideal relationship between the area covered by a water surface and the amount of cooling generated by this water body. Blue space and several summer temperature derivatives showed coefficients greater than 0.90, indicating a widespread thermal benefit during the day and night, and even during peak heatwave conditions provided by this surface type. However, while effective to reduce high temperatures in summer, blue space had no effect on minimum summer (mean, daytime and nighttime) temperatures (Figure 9).

4.2 EFFECT OF SURFACE TYPE ON SUMMER TEMPERATURESWhen analysing the entire near-surface air temperature dataset, spanning from December 2018 to February 2019 and across 102 locations, some clear trends emerged. Put simply, near-surface air temperatures warmed when the area taken up by roads, carparks and buildings increased. In contrast, open space and tree canopy helped cooling near-surface air temperatures.

A more refined view shows that the relative area covered by hard surfaces (e.g. roads, car parks and pathways) had a stronger effect on warming mean summer temperatures up (r = +0.71, p <0.001) compared to the area covered by buildings (r = +0.46, p = 0.001). The significant warming effect of these surface types is remarkable, because as shown in Figure 7 their area was on average less than 25% of the 50 m radius circle. However, the effect of warming was not increased when both surface types were combined. Hard surfaces and buildings had a marked impact on mean and minimum nighttime temperatures but increasing their area did not have any meaningful effect on maximum temperatures, either across the entire summer or during day- and nighttime (Figure 9).

Vegetated open space did not contribute to either warming or cooling of summer temperatures, whereas unvegetated open space did. Similar to tree canopy, the strongest cooling effect from unvegetated open space was detected for mean summer temperatures (r = -0.68, p <0.001), minimum summer temperatures (r = -0.66, p <0.001), mean summer nighttime (r = -0.70, p <0.001) and minimum summer nighttime temperatures (r = -0.65, p <0.001). Tree canopies were most effective to cool maximum nighttime temperatures in summer (r = -0.67, p <0.001), while they were less effective in reducing mean nighttime and mean summer temperatures (both r = -0.49, p <0.001). Increasing tree canopy had no measurable effect on reducing either maximum summer or maximum daytime temperatures. These extreme temperatures represent peak heatwave events where cooling from tree canopies did not provide any detectible cooling effects.

FIGURE 9: Colour-coded correlation matrix depicting if 11 surface types (single and combined) contributed to warming (red) or cooling (blue) of near-surface air temperatures during the summer season 2018/19. A greater influence, meaning a stronger correlation between the independent and dependent variable, of the surface type on warming or cooling is signified by the intensity of the colour. The grey colour indicates that a surface type had no influence on the temperature derivative. Daytime represents the interval between 10:00 and 18:00, nighttime was between 21:00 and 05:00 the following day.



21

FIGURE 10: Colour-coded correlation matrix depicting how strongly 11 surface types (single and combined) influence near-surface air temperatures during the summer season 2018/19. The strength of the relationship is indicated by the p-value of the Pearson correlation between the independent and dependent variables. This value is colour-coded from green (highly significant relationship with p ≤ 0.001) to red (no significant relationship with p > 0.05). Daytime represents the interval between 10:00 and 18:00, nighttime was between 21:00 and 05:00 the following day.

While Figure 9 can be used to ascertain the direction of the relationship (cooling or warming) among a single or composite surface type and its impact on summer temperature derivatives, Figure 10 provides an overview of the strength of these relationships. A colour code similar to a traffic light was applied to provide a simple representation of strong (green), weak (yellow) and missing (red) statistical relationships among the independent and dependent variables.

The aforementioned lack of effects on warming or cooling by vegetated open space is clearly visible by the number of red cells for this surface category in Figure 10. This effect is further enforced by the near identical outcome when assessing the relationship between summer temperature derivatives and the composite surface type category of vegetated surfaces and tree canopy (Figure 10). We suggest that this observation is largely driven by the disconnect between the air closely above vegetated ground surfaces and the atmospheric layer where cooling is actually influenced by air movement and transpirative cooling from tree canopies. It is well known that green lawn and other ground vegetation does not contribute much to air cooling in urban space. The larger impact of hard surfaces compared to buildings as drivers for warming in urban space is visible by the high number of green cells in Figure 10 for hard surfaces (n=7) and a lower number of green cells for buildings (n=5).



area of unvegetated open space resulted in higher warming during the morning, while a larger area of hard surfaces and buildings resulted in lowering the air temperature differentials. The reverse effect was captured for the nighttime interval, where unvegetated open space allowed for fast cooling, while hard surfaces and buildings warmed air temperatures.

The opposite trends of warming and cooling air masses by the different surface types can be explained by the thermal mass contained in the different surface types. Bare soil has a low thermal mass, resulting in fast absorption and release of solar energy. This surface type will heat up quickly during the morning and release stored energy as sensible heat into the surrounding environment, resulting in warming air temperatures. Asphalt, concrete or bricks have a higher thermal mass that takes longer to heat up in the morning, where solar energy is absorbed and not re-radiated into the surrounding environment. Thus, a high cover of these materials results in lower temperature differentials in the morning. However, while the low thermal mass of bare soil means that after sunset this surface type cools down quickly, the materials with high thermal mass will continue to re-radiate stored solar energy whereby they continue to warm the surrounding air during the night. Another textbook example for the origin of Urban Heat Island effects captured by these analyses.

observation that impervious surfaces (hard surfaces and buildings) have a very high impact on warming summer temperatures at night, even though they are representing only a small fraction of surface area. This effect is a textbook illustration of the Urban Heat Island effect at microscale.

During both months, no individual or combined surface type had a cooling or warming effect on daytime temperatures. This observation provides further evidence that significant effects on mean summer or mean monthly temperatures are driven to a large proportion by the cooling or warming influence of different surface types during the night. Highly significant cooling effects by unvegetated open space (January: r = -0.70, February: r = -0.72) and warming effects by hard surfaces from roads and carparks (January: r = +0.73, February: r = +0.68) were indeed present for mean nighttime temperatures during both months (Figure 11). Similar to temperatures for the entire summer, none of the surface types, except blue space had a statistically significant influence on maximum air temperatures in January or February.

It is worth pointing out that the analyses of temperature differentials between 06:00 and 10:00 (i.e., heating up in the morning) as well as 20:00 to 00:00 (i.e., cooling down during the night) captured thermodynamic processes well. As depicted in Figure 11, an increasing

4.3 EFFECT OF SURFACE TYPE ON JANUARY AND FEBRUARY TEMPERATURESJanuary was the hottest month in the summer of 2018/19 across the LGA of Campbelltown. Mean air temperature during January 2019 was 25.2°C. Mean daytime and nighttime temperatures during this month were 29.9 °C and 21.7°C. Air temperatures were notably lower in February, where mean monthly, daytime and nighttime temperatures were 22.6°C, 27.2, and 19.2°C respectively. Because of these large differences in cumulative heat load between January and February 2019, the capacity of surface types to cool or warm air temperatures was assessed for each month individually.

The Pearson correlation matrix in Figure 11 indicates a pronounced cooling capacity of unvegetated open space. These relationships are slightly weaker for mean temperatures in January (r = -0.63) compared to those in February (r = -0.70). The higher overall heat loads during January resulted in a stronger warming effect of roads and carparks on mean temperatures in January (r = +0.71) compared to February (r = +0.68). Notably, one of the highest Pearson correlation coefficients for all 308 tested models was detected for the influence of combined hard surfaces and buildings on warming up mean January nighttime temperatures (r = +0.75) (Figure 11). This result underpins the previous



23

The Pearson correlation coefficients for blue space were relatively high, especially for mean January (r = -0.91) and maximum January (r = -0.94) temperatures (Figure 11). As depicted in Figure 12, both relationships were highly significant, yet must be interpreted with caution due to the relatively low number of sites with blue space (n = 12). In contrast, the cooling capacity of open spaces (vegetated plus non-vegetated) and tree canopy was both strong (as indicated in Figure 11) and highly significant (Figure 12). Nearly all 102 sites were included in these combined categories, providing very high confidence in the cooling capacity of open spaces and tree canopy.

FIGURE 11: Colour-coded correlation matrix depicting if 11 surface types (single and combined) contributed to warming (red) or cooling (blue) of near-surface air temperatures during January and February 2019. A greater influence, meaning a stronger correlation between the independent and dependent variable, of the surface type on warming or cooling is signified by the intensity of the colour. The grey colour indicates that a surface type had no influence on the temperature derivative. Daytime represents the interval between 10:00 and 18:00, nighttime was between 21:00 and 05:00 the following day.



FIGURE 12: Colour-coded Pearson correlation matrix depicting how strongly 11 surface types (single and combined) influence near-surface air temperatures during the months of January and February 2019. The strength of the relationship is indicated by the p-value of the Pearson correlation between the independent and dependent variables. This value is colour-coded from green (highly significant relationship with p ≤ 0.001) to red (no significant relationship with p > 0.05). Daytime represents the interval between 10:00 and 18:00, nighttime was between 21:00 and 05:00 the following day. The processes of heating up and cooling down were captured by calculating the difference in air temperatures between 06:00 and 10:00 (heating up in the morning) and between 20:00 and 00:00 (cooling down in the night).



25

FIGURE 13: Relationships between a range of 11 individual and combined surface types and peak heatwave air temperatures during 5 individual days in December 2018 and January 2019. The top panel shows the colour-coded Pearson correlation matrix depicting how surface types contributed to warming (red) or cooling (blue) of near-surface air temperatures during the heatwaves. A greater influence of the surface type on warming or cooling is signified by the intensity of the colour. The grey colour indicates that a surface type had no influence on the temperature derivative. The bottom panel depicts the colour-coded Pearson correlation matrix for the strength of the relationship between surface types and heatwave temperatures. The strength is indicated by colour gradient from green (highly significant relationship with p ≤ 0.001) to red (no significant relationship with p > 0.05).

4.4 EFFECT OF SURFACE TYPE ON HEATWAVE CONDITIONSThe influence of any individual or combined surface type on peak heatwave conditions were generally very weak, although some general trends can be identified. This is because we have very limited control in mitigating temperatures during heat wave conditions when large masses of warm air move eastward from inland Australia. Hard surfaces and buildings generally had a warming effect, while unvegetated open space had a cooling effect. During peak heatwave conditions on 27 January 2019 air temperature across the LGA was 38°C, with maximum temperatures above 43°C (Site 25 at Holsworthy). Warming and cooling impacts on site-specific microclimates by hard surfaces and unvegetated open paces showed the strongest Pearson correlation factors and although relatively low (unvegetated open space: r = -0.48; hard surfaces: r = +0.41), these relationships were highly significant (p <0.001) (Figure 13).

The cooling effect of blue space on maximum air temperatures is clearly captured when selecting for individual extremely hot days. Except on 5 January 2019, increasingly large water surfaces helped lower peak heatwave temperatures. The remaining four Pearson correlation coefficients were greater than -0.8 for blue space during heatwaves.

Increasing tree canopy cover had a significant cooling effect on mean, minimum and nighttime summer temperatures. This cooling effect was also captured by analysing air temperatures for January and February. However, as shown in Figure 13, tree canopy cover did not have any significant cooling effect during four of the five heatwave events. Even when a statistically significant cooling effect (p = 0.0002) was detected on 5 January 2019, the relationship between tree canopy cover and air temperature was not very strong (r = -0.37).



FIGURE 14: Relationships between three surface types and three temperature derivatives calculated for the entire summer season of 2018/19. Lines show linear relationships, grey areas show 95% confidence intervals and the R2 value indicates the strength of the linear relationship. All relationships for mean summer and maximum summer nighttime temperatures were highly significant (p <0.0001). Those for maximum summer daytime temperatures were only significant for hard surfaces (p = 0.01). Note: the relative area covered by vegetated and unvegetated open space and tree canopy can exceed 100% at individual sites. This effect is due to the treatment of canopy cover as separate (‘aboveground’) layer; adding ground and aboveground layers can result in a value greater than 100%.

4.5 INDIVIDUAL WARMING AND COOLING EFFECTS To support the development of improved planning strategies that aim to mitigate urban heat, it is necessary to further examine the complex relationships between individual and combined surface types with a range of temperature derivatives. As previously shown, increasing the area covered by a single surface type can result in cooling or warming of air temperatures. However, changing the proportional cover of a surface type or tree canopy cover will have different effects on different temperature derivatives.

The analyses reveal that increasing the coverage of hard surfaces will result in warming of mean summer temperatures, and the greater cover of hard surfaces will also lead to warmer summer nights (Figure 14). Yet, increasing hard surfaces will have very little effect on maximum summer daytime temperatures, which also applies to any increases in tree canopy cover or the provision of open space. Maximum summer daytime

temperatures are recorded during peak heatwave conditions. During these times, super-heated air masses originating from the centre of Australia are largely responsible for extreme air temperatures. The energy captured in these air masses far exceeds the capacity of hard surfaces to further warm, or tree canopy and open spaces to cool down the extreme air temperatures. Analyses provided in figure 14 show, however, increasing the area of tree canopy cover and open space will provide markedly cooler nights.

Greater coverage of open space (vegetated or unvegetated) resulted in cooling effects, while hard surfaces and buildings caused warming during summer and particularly during nighttime. However, during periods of extreme summer heat the warming or cooling effect of these surface types on air temperatures diminished (Figure 15). While an increase in the cover of hard surfaces and buildings or open space from 20% to 50% would warm mean summer temperatures by 0.79°C (Figure 16), it would only increase

maximum air temperatures by 0.29°C. Similarly, increasing the area covered by open space (combination of vegetated and unvegetated surfaces) from 20% to 50% would reduce mean summer temperatures by 0.49°C (Figure 16), but only lower maximum air temperatures by 0.35°C (Figure 15).

Figure 15 also shows that very high air temperatures were recorded at sites with a high percentage of open space and a low percentage of hard surfaces and buildings. These sites were predominately located in native bushland and open forests around the eastern part of the LGA where open space heated up quickly during the day without absorbing much solar radiation. While these locations heated up most during the day, they also cooled down most during the night. This phenomenon was reported in detail by Pfautsch and Rouillard (2019).

Previous analyses highlighted the role of blue space in cooling down hot and extreme summer air temperatures. Other surface types had much less or no cooling or warming



27

FIGURE 15: Relationships between four individual and combined surface types and extreme temperatures, calculated for the entire summer 2018/19 (upper panels) and the peak heatwave conditions during 31 January 2019 (lower panels). Lines show linear relationships, grey areas show 95% confidence intervals and the R2 value indicates the strength of the linear relationship. All relationships, except those for tree canopy, were statistically significant (p <0.01).

influence on extreme temperatures. A detailed look at the relationship of blue space with extreme temperatures reveals that two specific sites dominated the trend (Figure 15). These specific sites lacked hard surfaces and buildings and were located along the Nepean River at Spring Farm (Site 97) and O’hares Creek in the Dharawal National Park (Site 77). The high percentage of open ground space and tree canopy cover combined with evaporative cooling from water bodies would have resulted in the generally cooler microclimate at these two sites. If these two sites would be excluded from the analyses, no significant cooling could be detected for blue space. This highlights how only the combination of a range of factors may lead to a marked reduction of peak summer heat. Nonetheless, the cooling effect of large water bodies is well known and generally applies to selected locations around the LGA of Campbelltown.



well as to mean daytime and nighttime temperatures calculated for the months of January and February 2019. These relationships demonstrate that increasing either unvegetated or vegetated open space from 10% to 50% cover will lower mean summer daytime temperatures by 0.51°C, and reduce mean summer nighttime temperatures by 0.80°C (Table 3).

Reductions in daytime and nighttime temperature provided by increasing open space were small compared to the increases in air temperatures as a result of expanding the area covered by hard surfaces. When increasing the coverage of hard surfaces from 10% to 40%, mean daytime temperatures during summer will increase 0.73°C and nighttime temperatures go up by 1.27°C (Table 3). Due to the steeper slope and intercept of the regressions for hard surfaces the warming effect of hard surfaces largely exceeded any

cooling benefits of open space at the high end of relative surface cover. This means that at a hypothetical site with equal cover of open space and hard surfaces, temperatures will increase. As a result of the steeper slope and intercept combined with the greater influence of hard surfaces on nighttime warming indicates that effective cooling at sites that contain both surface types can only be achieved once the ratio of open space to hard surfaces is 2:1 or greater.

As mentioned earlier, the impact of daytime cooling by increased tree canopy cover will be relatively small. Analyses of the data indicate that there was no effect of tree canopy on either mean or maximum daytime temperatures and the relationships with daytime temperatures in January and February 2019 were considerably weak. However, the cooling effect through canopy cover on nighttime air temperatures was

FIGURE 16: Predicted changes in mean summer air temperature along a gradient of different surface cover types and tree canopy. Values above the zero line indicate warming, those below cooling. Data for temperature changes from 20% to 50% relative surface cover of hard surface and open space (sum of vegetated and unvegetated surfaces) are shown as example for the magnitude of warming and cooling effects.

These individual relationships can be used to determine if increasing or decreasing the area of a specific surface type will result in warming or cooling air temperature in summer. While increasing hard surfaces from 20% to 50% will lead to warming mean summer temperatures by 0.76°C, increasing the same amount of surface area covered by open space (both vegetated and unvegetated surface types) will lower mean summer temperatures by 0.49°C (Figure 16). The reverse effect can be expected when decreasing hard surfaces (resulting in cooler summer temperatures) or open space (resulting in warmer summer temperatures).

Assessing these relationships further indicates that the impact of warming from hard surfaces like roads and carparks is greater compared to the warming effect of buildings (Figure 16). On average, 16% of the area within each of the 102 measurement circles was covered by hard surfaces, yet only by 7% of the area was covered with buildings, which can explain the greater warming effect of hard surfaces. In addition, the study assessed near-surface air temperatures, 2-3 m above the surface. Radiant heat from planar roads and other hard surfaces could have heated air close to the ground more compared to vertical walls from buildings.

The overall greater cooling effect of open space compared to tree canopy cover (Figure 16) can be explained by the ‘blanket function’ of tree crowns and the retention of warmed air during the night. During the evening and into the night open space will lose heat absorbed during the day quickly by unobstructed convection into the surrounding atmosphere. If a tree canopy is established above open space, it will trap the warmed air – similar to a blanket – underneath the canopy and thus keep air temperatures slightly warmer.

The importance of planning and managing urban space to mitigate heat not only during the day, but also during the night becomes obvious when assessing the observed warming and cooling trends for the two surface types and tree canopy (Table 3). Commonly, cooling benefits generated by increasing the area covered by open space are greater during the night compared to those during the day. This applies to mean and maximum summer temperatures, as



29

consistent and highly significant. Cooling benefits during warm or hot nights provided by tree canopy cover exceeded even those provided by open spaces when their respective surface cover was equal. This result highlights the importance of tree canopy cover in managing thermal comfort of local communities during hot nights.

SURFACE TYPE

SURFACE COVER (%)

MEAN SUMMER DAYTIME

TEMP.

MEAN SUMMER

NIGHTTIME TEMP.

MAX. SUMMER DAYTIME

TEMP.

MAX. SUMMER

NIGHTTIME TEMP.

MEAN JANUARY DAYTIME

TEMP.

MEAN JANUARY

NIGHTTIME TEMP.

MEAN FEBRUARY DAYTIME

TEMP.

MEAN FEBRUARY NIGHTTIME

TEMP.Open space 0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

10 -0.13 -0.20 -0.18 -0.19 -0.14 -0.19 -0.13 -0.23

20 -0.26 -0.40 -0.35 -0.38 -0.27 -0.37 -0.26 -0.47

30 -0.39 -0.60 -0.53 -0.58 -0.41 -0.56 -0.40 -0.70

40 -0.51 -0.80 -0.70 -0.77 -0.55 -0.75 -0.53 -0.94

50 -0.64 -1.00 -0.88 -0.96 -0.68 -0.93 -0.66 -1.17

60 -0.77 -1.20 -1.05 -1.15 -0.82 -1.12 -0.79 -1.41

70 -0.90 -1.40 -1.23 -1.35 -0.95 -1.31 -0.93 -1.64

80 -1.03 -1.60 -1.40 -1.54 -1.09 -1.49 -1.06 -1.87

90 -1.16 -1.80 -1.58 -1.73 -1.23 -1.68 -1.19 -2.11

100 -1.29 -2.00 -1.76 -1.92 -1.36 -1.87 -1.32 -2.34

Hard surfaces

0 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00

10 +0.18 +0.32 +0.22 +0.30 +0.19 +0.30 +0.18 +0.36

20 +0.37 +0.64 +0.43 +0.59 +0.37 +0.59 +0.36 +0.72

30 +0.55 +0.96 +0.65 +0.89 +0.56 +0.89 +0.54 +1.08

40 +0.73 +1.27 +0.86 +1.18 +0.74 +1.19 +0.72 +1.44

50 +0.91 +1.59 +1.08 +1.48 +0.93 +1.48 +0.90 +1.80

60 +1.10 +1.91 +1.30 +1.77 +1.11 +1.78 +1.08 +2.17

70 +1.28 +2.23 +1.51 +2.07 +1.30 +2.07 +1.26 +2.53

80 +1.46 +2.55 +1.73 +2.36 +1.48 +2.37 +1.44 +2.89

90 +1.64 +2.87 +1.95 +2.66 +1.67 +2.67 +1.62 +3.25

100 +1.83 +3.19 +2.16 +2.95 +1.86 +2.96 +1.81 +3.61

Tree canopy 0 - 0.00 - 0.00 0.00 0.00 0.00 0.00

10 - -0.14 - -0.23 -0.06 -0.15 -0.06 -0.19

20 - -0.29 - -0.45 -0.12 -0.30 -0.13 -0.38

30 - -0.43 - -0.68 -0.17 -0.44 -0.19 -0.58

40 - -0.57 - -0.91 -0.23 -0.59 -0.26 -0.77

50 - -0.72 - -1.13 -0.29 -0.74 -0.32 -0.96

60 - -0.86 - -1.36 -0.35 -0.89 -0.39 -1.15

70 - -1.00 - -1.59 -0.40 -1.04 -0.45 -1.34

80 - -1.14 - -1.82 -0.46 -1.18 -0.51 -1.54

90 - -1.29 - -2.04 -0.52 -1.33 -0.58 -1.73

100 - -1.43 - -2.27 -0.58 -1.48 -0.64 -1.92

TABLE 3: Estimated change in air temperature due to a change in the area covered by open space (vegetated and non-vegetated ground surfaces combined), hard surfaces (like roads and carparks) and tree canopy cover. A positive sign indicates warming, a negative sign indicates cooling. Only data for statistically significant relationships (p <0.05) are shown. The table shows temperature changes in °C for four daytime (white columns) and four nighttime (grey columns) temperature derivatives. The values in the table must be read as ‘deviation from the air temperature at a site where the specific surface type is entirely absent’ (i.e., 0% cover provides zero deviation in air temperature).



frequently compared to suburbs where the proportion of open space and canopy cover is greater. A promising strategy to reduce the cover of hard surfaces and buildings is known as ‘smart density’. It promotes a combination of compact building structure and heat-adverse building design (Middel et al. 2014). When applying the principle findings of the present study, the term can be expanded to ‘heat-smart density’ that will increase resilience of local communities against heat by providing ample space for green and blue infrastructure, while medium- and high-rise buildings are designed to provide maximum shade on hard surfaces and adjacent buildings. Underground carparks would further reduce the area of hard surface that absorbs solar radiation during the day and emits sensible heat during the night.

To reduce the amount of heat contributed by hard surfaces and buildings to urban microclimates, and based on findings of the present report it is recommended that:

1. The area covered by hard surfaces is reduced wherever land use allows a transformation to open, pervious surface

2. Where such a transformation is not possible, strategies to reduce the exposure of hard surfaces to solar radiation are applied (e.g. shading)

3. Where shading of hard surfaces is not possible, strategies to increase the surface albedo (e.g., reflective surface coating) or change the surface material (e.g., replacing dark asphalt with light concrete or crushed/compacted sandstone) are applied

4. The area of new hard surfaces constructed during greenfield development is reduced to the absolute minimum

5. Mandate the implementation of recommendations 2, 3 and 4 for the approval of new greenfield development to mitigate heat from the get-go

5.1 REDUCE WARMINGIn agreement with international research (e.g., Guo et al. 2016, Jenerette et al. 2016, Morabito et al. 2016), the present work clearly identified the warming potential of typical grey urban infrastructure, such as roads, carparks, walkways, driveways and buildings. On average these hard surfaces covered only 23% of the surface space present at the 102 sites, but their impact on warming air temperatures was undeniable. Hard surfaces and buildings were the only two surface categories for which a persistent and significant warming effect on microclimates was detected. Increasing their area will lead to warmer mean and maximum summer air temperatures, with especially pronounced effects on nighttime air temperatures.

The strong warming of nighttime air temperature, as well as the slower heating of air temperatures during the morning at locations where hard surfaces and buildings covered a large proportion of the surface illustrates the well-known Urban Heat Island Effect at microscale. It is well documented that pronounced Urban Heat Island Effects, and especially nighttime heat during heatwave events, lead to increased hospital admissions and rising mortality rates (e.g., Åström et al. 2011, Zhang et al. 2012). Hence, strategies that aim at reducing nighttime heat have a positive effect on local communities, where those most vulnerable to heat will experience the greatest benefits (Harlan et al., 2006). This group includes the elderly, very young children, immobile and chronically sick people and those that do not have the financial means to pay for air conditioning.

Suburbs developed in the past decade across in the western part of the Campbelltown LGA and other regions in Western Sydney have an alarmingly low cover of open space and green infrastructure but a high cover of hard surfaces and buildings. Data presented here evidences that this type of suburban development will leave local communities exposed to high and extreme heat loads more

This research project assessed if and how different surface types and also tree canopy cover influence air temperatures across the LGA of Campbelltown City in the summer of 2018/19. More than 1.2 million measurements of near-surface air temperature were used to investigate microclimate effects at 102 sites. General climate trends and distinctive heat-related patterns across these sites were described in an earlier report (Pfautsch and Rouillard 2019).

5. RECOMMENDATIONS



31

Pacific Oceans. These systems are immune to the potential impact of artificial water bodies in Western Sydney.

As suggested by the Sydney Water Corporation (2017), other techniques and technologies that involve water for cooling seem more feasible. These include implementing water-based technologies into local planning and enforce water sensitive urban design in Development Control Plans (DCPs) and Local Environmental Plans (LEPs). Further, construction of small-scale water bodies and protection and revitalisation of existing local creek lines can become mandatory elements for the approval of greenfield development. Responsible and progressive urban planning and assessments of environmental impacts should have a strong emphasis on protecting and recycling local water sources with the goal to provide water for a range of local cooling strategies.

5.4 REGULATING URBAN TEMPERATURES WITH TREE CANOPYAnalyses presented here document a weak cooling capacity of tree canopies. Tree canopy cover did not reduce mean summer or maximum daytime air temperatures. However, significant cooling was provided by tree canopy cover during the night. At the same time, the analyses also documented a ‘blanket effect’ of tree canopies that traps warm air during the night. Both processes are not in conflict with each other and can be explained by differences in effect size.

Mixing of air during the day will result in very similar air temperatures underneath trees compared to surrounding unshaded space. The mixing effect explains why expanding tree canopy cover generates little cooling benefits on air temperatures during the day. However, canopy shade reduces the amount of heat absorbed by surfaces during the day, which in turn will result in cooler air temperatures at the beginning of the night and less emission of heat during the night. This effect is relatively large in size and it is at night when the cooling benefits of canopy shade become apparent. A smaller residual effect of tree canopy cover trapping warm air during the night remains,

Based on the omnipresent cooling capacity of open spaces in summer, it is recommended that:1. Open space is retained wherever possible

2. Unused or obsolete areas covered by hard surfaces are transformed to open space

3. Additional tree plantings in open space are concentrated to the edges of the area, where additional thermal benefits can be generated by shading adjoining hard surfaces like roads, walkways and carparks

4. Policies are developed that mandate the minimum proportion of 2:1 for open space and hard surfaces in greenfield development

5.3 THE ROLE OF BLUE SPACEAlthough the capacity of blue space to cool urban microclimates has been documented in other research reports (e.g., Sun et al. 2012, Sun and Chen 2012, Hongyu et al. 2016), including for Western Sydney (Sydney Water Corporation 2017), care must be taken when interpreting the results around blue space in this report. The strong cooling capacity of water bodies was driven by two locations out of a total of 12. Both locations were near large water bodies that represent an exception, rather than a common feature in the southwestern part of the Sydney Basin.

Enhancing the cooling benefits of blue space in the area of Campbelltown would require the constructions of large artificial water bodies. The chance to realise such infrastructure projects seems very low for a number of reasons. First, capital investment to acquire the necessary land may be cost prohibitive. Second, the amount of water necessary to fill and maintain large artificial water bodies may not be available. Third, environmental conditions that precede extreme heat and heatwaves are prolonged periods of low rainfall and hot surface soils, which would put pressure on artificial water bodies through high rates of evaporation and the lack of recharge. And fourth, heatwaves that impact Western Sydney have their origin in central Australia. Synoptic patterns over this vast landmass depend on conditions generated by climate systems over the Arctic, Indian and

5.2 INCREASE COOLINGIn contrast to the warming effect of hard surfaces and buildings, a ubiquitous cooling benefit was documented especially for unvegetated open space. However, the contribution to cooling differed markedly if a surface was covered by vegetation or not. The separate analyses of thermal benefits from these surface types revealed that relatively bare surfaces provided far greater cooling benefits compared to surfaces covered with green lawns and other ground vegetation. This seems counterintuitive and requires further assessment.

For the present work, aerial images were used to classify and measure surface types at each site. High resolution imagery was available for late December 2018 and late March 2019. Both dates were outside the core research interval of January and February. While imagery from December shows widespread green cover of open spaces, the March imagery shows that the same surfaces in yellow and brown colours, indicating that ground cover vegetation had suffered from the hot and dry conditions starting in late December (i.e., Boxing Day heatwave 2018). These conditions lasted until the beginning of March. The March image was used for the present analyses, because ground vegetation in this image was more representative for dry, hot summer conditions. This will inevitably have resulted in assigning the ‘unvegetated surface’ category to many grassed or otherwise vegetated areas.

Analyses are provided for both individual surface types and in most cases also for their combined effect on near-surface air temperatures (or in combination with tree canopy). These analyses show outright cooling benefits of what was termed ‘open space’ throughout the present report. Open space was shown to reduce mean and maximum summer temperatures and also help cool down daytime and nighttime temperatures. Open space also helped in cooling down these temperatures more compared to tree canopy cover. This result underlines the importance of retaining non-treed open space in urban landscapes, as these spaces maximise cooling benefits during the night where stored surface heat can escape unobstructed into the lower atmosphere.

5. RECOMMENDATIONS



5.5 LOW CAPACITY TO REDUCE HEATWAVE TEMPERATURESLastly, the present work provides evidence that the capacity to lower peak heatwave temperatures by modifying the configuration of urban surface area and tree canopy cover is very limited. Expanding the area of tree canopy did not result in a systematic cooling effect and was shown to be ineffective during four of five days where air temperatures at many of the 102 locations were above 40°C. Cooling benefits during extreme heat were only provided by blue space. The cooling capacity of large water bodies has been demonstrated in modelling studies. However, constructing large artificial water bodies, designed to lower peak heatwave temperatures is not practical for a number of reasons described above.

Recent climate predictions for south-eastern Australia point towards increasing summer heat and more frequent, lasting and extreme heatwaves. It is recommended that other means are developed to protect local populations and economies from peak heatwave conditions. These means should include Heat Emergency Plans that detail trigger points, and mechanisms to provide respite to the community (Zhang et al. 2012). While such plans represent a vital strategy to cope with increasing risks of extreme heat, they are beyond the scope of the present work.

signified by the greater cooling effect of open space where canopy cover is low or absent (see Figure 16). This clarifies how tree canopy cover provides both, a larger cooling effect by shading urban space during the day that keeps nighttime air temperatures lower due to reduced emission of stored heat and a smaller warming effect through trapping warm air at night.

Based on the findings of the present work and other published research it is recommended to:1. Tighten existing guidelines and policies to

slow current rates of loss of mature urban trees that provide the largest cooling benefits through surface shading

2. Tree removal permits become mandatory when removing any trees

3. Focus active management on existing younger trees to support crown development and canopy expansion

4. Introduce and maintain additional trees into urban space, especially where these trees will shade hard surfaces

5. Move tree plantings towards the edges of open space that can provide valuable nighttime cooling (and tree canopy cover would trap warm air)

The present work has also highlighted that the cooling effect of open space is greater compared to that of tree canopies, providing an incentive to look elsewhere for cooling opportunities that unlike tree crowns do not require years or decades to unfold their potential to mitigate urban heat (e.g., green trellises, artificial shade structures). Still, urban trees deliver many benefits other than cooling, which makes them the most valuable urban green infrastructure asset in urban landscapes.



33



Many options are available to plan and develop the cities and towns of tomorrow. Based on the present analyses, ‘heat-smart density’ seems a promising candidate for Western Sydney. In this fast-growing region of the Sydney Basin, development needs to provide homes for many thousands of new residents. At the same time the required buildings must be less dependent on electricity to provide thermal comfort for their residents and the suburban spaces must be better designed to help local communities to thrive.

The time to build these settlements is now. This report helps to do it.

Here, estimates of cooling and warming of microclimates in urban space were calculated based on the relative cover of five common surface types and tree canopy cover. These estimates can be used to develop improved policy and planning frameworks that incorporate the thermal effects of surface cover on urban microclimates.

In the past decade, heat-smart urban design was largely ignored across Western Sydney. Yet, heat-smart urban design is becoming increasingly important to mitigate the effects of warming summers.

6. CONCLUSION



35

Environmental Hazards 6, 39-49.

Song J., Du S., Feng X., Guo L. (2014) The relationships between landscape compositions and land surface temperature: quantifying their resolution sensitivity with spatial regression models. Landscape and Urban Planning 123, 145-157.

Sun R.H., Chen A.L., Chen D.L., Lu Y.H. (2012) Cooling effects of wetlands in an urban region: the case of Beijing. Ecological Indicators 20, 57-64.

Sun R.H., Chen L.D. (2012) How can urban water bodies be designed for climate adaptation? Landscape and Urban Planning 105, 27-33.

Sydney Water Corporation (2017) Cooling Western Sydney – a strategic study on the role of water in mitigating urban heat in Western Sydney. Sydney Water Corporation, 24 p.

Wang Y., Berardi U., Akbari, H. (2016a) Comparing the effects of urban heat island mitigation strategies for Toronto, Canada. Energy and Buildings 114, 2-19.

Wang J., Qinming Z., Guo H., Jin Z. (2016b) Characterizing the spatial dynalics of land surface temperature-impervious surface fraction relationship. International Journal of Applied Observation and Geoinformation 45, 55-65.

Ward K., Lauf S., Kleinschmit B., Endlicher W. (2016) Heat waves and urban heat islands in Europe: A review of relevant drivers. Science of the Total Environment 569-570, 527-539.

Zhang K., Rood R.B., Michailidis G., Oswald E.M., Schwartz J.D., Zanobetti A., Ebi K.L., O’Neill M.S. (2012) Comparing exposure metrics for classifying ‘dangerous heat’ in heat wave and health warning systems. Environment International 46, 23-29.

Zhao, L., Lee, X., Smith, R.B., Oleson, K., 2014. Strong contributions of local background climate to urban heat islands. Nature 511, 216–219.

Ziter, C.D., Pedersen, E.J., Kucharik, C.J., Turner, M.G., 2019. Scale-dependent interactions between tree canopy cover and impervious surfaces reduce daytime urban heat during summer. Proceedings of the National Academy of Science of the United States 116, 7575–7580.

impacts on pedestrian-level climate. Urban Climate 32, 100590.

Lowe S.A. (2016) An energy and mortality impact assessment of the urban heat island in the US. Environmental Impact Assessments 56, 139-144.

Middel A., Häb K., Brazel A.J., Martin C.A., Guhathakurta S. (2014) Impact of urban form and design on mid-afternoon microclimate in Phoenix Local Climate Zones. Landscape and Urban Planning 122, 16-28.

Morabito M., Crisci A., Gioli B., Gualtieri G., Toscano P., Di Stefano V., Orlandini S., Gensini, G.F. (2015) Urban-hazard risk analysis: mapping of heat-related risks in the elderly in major Italian cities. PLoS One 10, e0127277.

Morabito M., Crisci A., Messeri A., Orlandini S., Raschi A., Maracchi G., Munafò M. (2016) The impact of built-up surfaces on land surface temperatures in Italian urban areas. Science of the Total Environment 551-552, 317-326.

Oke T., 1973. City size and the urban heat island. Atmosphere and Environment 1967, 769-779.

Pfautsch, S., Rouillard, S. 2019. Benchmarking heat across Campbelltown, New South Wales. Western Sydney University, 60p.

Roth M., Chow W.T.L. (2012) A historical review and assessment of urban heat island research in Singapore. Singapore Journal of Tropical Geography 33, 381-397

Santamouris M. (2015) Analyzing the heat island magnitude and characteristics in one hundred Asian and Australian cities and regions. Science of the Total Environment 512-513, 582-598.

Schwarz N., Schlink U., Franck U., Großmann K. (2012) Relationship of land surface and air temperatures and its implications for quantifying urban heat island indicators – an application for the city of Leipzig (Germany). Ecological Indicators 18, 693-704.

Skelhorn C., Lindley S., Levermore G. (2014) The impact of vegetation types on air and surface temperatures in a temperate city: a final scale assessment in Manchester, UK. Landscape and Urban Planning 121, 129-140.

Solecki W.D., Rosenzweig C., Parshall L., Pope G., Clark M., Cox J., Wiencke M. (2011) Mitigation of the heat island effect in urban New Jersey.

Åström D.O., Forsberg B., Rocklöv J. (2011) Heat wave impact on morbidity and mortality in the elderly population: a review of recent studies. Maturitas 69, 99-105.

Coutts A.M., Beringer J., Tapper, N.J. (2007) Impact of increasing urban density on local climate: spatial and temporal variations in the surface energy balance in Melbourne, Australia. Journal of Applied Meteorology and Climatology 46, 477-493.

Duncan J.M.A., Boruff B., Saunders A., Sun Q., Hurley J., Amati M. (2019) Turning down the heat: an enhanced understanding of the relationship between urban vegetation and surface temperature at the city scale. Science of the Total Environment 656, 118-128.

Estoque R.C., Murayama Y., Myint S.W. (2017) Effects of landscape composition and pattern on land surface temperature: an urban heat island study in the megacities of Southeast Asia. Science of the Total Environment 577, 349-359.

European Commission (2012) Guidelines on best practice to limit, mitigate or compensate soil sealing. European Commission, Brussels.

Guo G., Zhuo X., Wu Z., Xiao R., Chen Y. (2016) Characterizing the impact of urban morphology heterogeneity on land surface temperature in Guangzhou, China. Environmental Modelling and Software 84, 427-439.

Harlan S.L., Brazel A.J., Prashad L., Stefanov W.L., Larsen L. (2006) Neighborhood microclimates and vulnerability to heat stress. Social Science and Medicine 63, 2847-2863.

Hung T., Daisuke U., Shiro O., Yoshifumi, Y. (2006) Assessment with satellite data of the urban heat island effects in Asian mega cities. International Journal of Applied Earth Observations 8, 34-48.

Jenerette G.D., Harlan S.L., Buyantuev A., Stefanov W.L., Decelt-Barreto J., Ruddell B.L., Myint S.W., Kaplan S., Li X. (2016) Micro-scale urban surface temperatures are related to land-cover features and residential heat related health impacts in Phoenix, AZ USA. Landscape Ecology 31, 745-760.

Krayenhoff, E.S., JKiang T., Christen A., Martilli A., Oke T.R., Bailey B.N., Nazarin N., Voogt J.A., Giometto M.G., Stastyn A., Crawford B.A. (2020) A multi-layer urban canopy meteorological model with trees (BEP-Street): Street tree

7. REFERENCES



On the following pages, all 102 project sites are listed. Site-specific information is provided. One urban heat logger was located in the centre of each circle (50 m radius). Polygons delineate each of five different surface types or tree canopy. Tree canopies were assessed as a second layer, resulting in overlap of polygons delineating tree canopy with polygons of other surface classes. Data for each surface type and tree canopy is provided.

8. DETAILS OF STUDY SITES

LOCATION 1Serial number S118081339-16

Latitude -33.9544

Longitude 150.8622

Street Camden Valley Way

Suburb Edmondson Park

Additional info Outside Endeavour sub-station

Tree species Eucalyptus tereticornis

Vegetated open space (%) 20.79

Unvegetated open space (%) 35.43

Roads/carparks (%) 32.85

Buildings (%) 8.21

Blue space (%) 0.00

Tree canopy cover (%) 20.08



37


Latitude -33.9560

Longitude 150.8805

Street Casula RFS compound

Suburb Casula

Additional info

Tree species Eucalyptus micorcorys




Buildings (%) 7.82

Blue space (%) 0.00





Latitude -33.9518

Longitude 150.9030

Street Camden Valley Way

Suburb Casula

Additional info Near Dan Murphey’s





Buildings (%) 12.93

Blue space (%) 0.57




39


Latitude 33° 58’ 7.002” S (coordinates were initially wrong at -33.9564)

Longitude 150° 55’ 4.002” E (coordinates were initially wrong at 150.9241)

Street Old number 4

Suburb Holsworthy

Additional info off Moorebank Ave, near gas house

Tree species Eucalyptus crebra




Buildings (%) 1.20

Blue space (%) 0.00




LOCATION 5 Serial number S118081335-19

Latitude -33.9720

Longitude 150.9329

Street Old number 5

Suburb Holsworthy

Additional info

Tree species Eucalyptus punctata




Buildings (%) 0.00

Blue space (%) 0.00




41


Latitude -33.9734

Longitude 150.8888

Street Hurlestone - First point

Suburb Glenfield

Additional info Just around the bend

Tree species Casuarina cunninghamiana




Buildings (%) 0.00

Blue space (%) 0.00





Latitude -33.9713

Longitude 150.8799

Street Hurlestone - furthest point

Suburb Glenfield

Additional info

Tree species African Olive




Buildings (%) 0.00

Blue space (%) 0.00




43


Latitude -33.9722

Longitude 150.8624

Street Campbelltown Road

Suburb Bardia

Additional info Near bus stop opposite Dahua land





Buildings (%) 0.17

Blue space (%) 0.00



Latitude -33.9713

Longitude 150.8799

Street Hurlestone - furthest point

Suburb Glenfield

Additional info





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -33.9761

Longitude 150.8486

Street Regional Conservation Area

Suburb Edmondson Park

Additional info





Buildings (%) 0.00

Blue space (%) 0.00




45


Latitude -33.9740

Longitude 150.8261

Street Denham Court Road

Suburb Denham Court

Additional info Downhill from Fox Valley Road intersection

Tree species Eucalyptus mollucana




Buildings (%) 0.00

Blue space (%) 0.00





Latitude -33.9799

Longitude 150.8559

Street Vevi Street

Suburb Bardia

Additional info Light pole

Tree species




Buildings (%) 39.28

Blue space (%) 0.00




47


Latitude -33.9823

Longitude 150.8522

Street Wootten Avenue

Suburb Bardia

Additional info Light pole just up from the corner of Zouch Road

Tree species




Buildings (%) 10.93

Blue space (%) 0.21




LOCATION 13Serial number 118081335-02

Latitude -33.9754

Longitude 150.8589

Street MacDonald Rd

Suburb Bardia

Additional info tree behind the old school

Tree species Lophostemon confertus




Buildings (%) 4.65

Blue space (%) 0.00




49

LOCATION 14 Serial number 118081339-20

Latitude -33.9739

Longitude 150.8152

Street Jamboree Av

Suburb Willowdale


Tree species




Buildings (%) 0.00

Blue space (%) 0.00





Latitude -33.9791

Longitude 150.8201

Street off Commissioners Dr

Suburb Willowdale

Additional info Light pole, no street name available

Tree species




Buildings (%) 27.14

Blue space (%) 0.00




51


Latitude -33.9886

Longitude 150.8144

Street Lovegrass Avenue

Suburb Willowdale

Additional info

Tree species Light pole




Buildings (%) 51.17

Blue space (%) 0.00





Latitude -33.9722

Longitude 150.8042

Street Heath Roads

Suburb Leppington

Additional info Near substation on powerline Leppington





Buildings (%) 1.64

Blue space (%) 0.00




53


Latitude -33.9926

Longitude 150.8076

Street St Saba College

Suburb Varroville

Additional info Serbian Orthodox





Buildings (%) 0.42

Blue space (%) 0.00





Latitude -33.9936

Longitude 150.8225

Street Varroville Reserve

Suburb Varroville

Additional info





Buildings (%) 0.00

Blue space (%) 0.00




55


Latitude -33.9914

Longitude 150.8439

Street Campbelltown Road

Suburb Denham Court

Additional info

Tree species Eucalyptus eugenoides




Buildings (%) 0.00

Blue space (%) 0.00




LOCATION 21

Serial number S118081349-06

Latitude -33.9978

Longitude 150.8655

Street Ingleburn CBD

Suburb Ingleburn

Additional info Outside bakery, near Michelle’s patisserie

Tree species Ulmus parvifolia




Buildings (%) 32.67

Blue space (%) 0.00




57


Latitude -33.9915

Longitude 150.8623

Street 10 York Road

Suburb Ingleburn

Additional info Industrial estate

Tree species Eucalyptus nichollii




Buildings (%) 29.37

Blue space (%) 0.00





Latitude -33.9908

Longitude 150.882

Street 3 Bunbury Road

Suburb Macquarie Fields

Additional info In front of private house

Tree species Eucalyptus robusta




Buildings (%) 21.05

Blue space (%) 0.00




59


Latitude -33.9910

Longitude 150.9004

Street Eucalyptus Drive

Suburb Macquarie Fields

Additional info Caley Park

Tree species Eucalyptus fibrosa




Buildings (%) 5.99

Blue space (%) 0.00





Latitude -33.9946

Longitude 150.9213

Street Old number 19

Suburb Holsworthy

Additional info Parallel to Macquarie Fields

Tree species Allocasuarina littoralis




Buildings (%) 0.00

Blue space (%) 0.00




61


Latitude -34.0109

Longitude 150.9187


Suburb Holsworthy

Additional info Parallel to number 21





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.0083

Longitude 150.9025


Suburb Holsworthy

Additional info Top of Long Point, just south of bend

Tree species Angophora costata




Buildings (%) 0.00

Blue space (%) 0.00




63


Latitude -34.0108

Longitude 150.8824

Street Corner of Bensley and Oxford Road

Suburb Ingleburn

Additional info





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.0104588

Longitude 150.8644258

Street Desmond Street

Suburb Ingleburn

Additional info Edna Reserve





Buildings (%) 0.00

Blue space (%) 0.00




65


Latitude -34.0063

Longitude 150.8447

Street Stennart Road

Suburb Ingleburn

Additional info





Buildings (%) 3.97

Blue space (%) 0.00





Latitude -34.0095

Longitude 150.8235

Street Corner of Spitfire Drive and St Andrews Road

Suburb Varroville

Additional info Opposite entry to Mt Carmel High School





Buildings (%) 0.00

Blue space (%) 0.00




67


Latitude -34.0112

Longitude 150.8015

Street Macarthur Grange Country Club

Suburb Varroville

Additional info Street tree at entrance

Tree species Corymbia maculata




Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.0165

Longitude 150.7748

Street Fairbank Road

Suburb Gledswood Hills

Additional info

Tree species Eucalyptus microcorys




Buildings (%) 21.24

Blue space (%) 0.47




69

LOCATION 34

Serial number S118081339-05

Latitude -34.0255

Longitude 150.7731

Street Kavanagh Street

Suburb Gregory Hills

Additional info





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.0276

Longitude 150.8052

Street Epping Forest Drive

Suburb Eschol Park

Additional info Vale Brooke Reserve, western side

Tree species Corymbia citriodora




Buildings (%) 18.14

Blue space (%) 0.89




71


Latitude -34.0290

Longitude 150.8234

Street Raby Road

Suburb Raby

Additional info Road reserve on ramp to M5





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.0285

Longitude 150.8425

Street Coronation Car Park

Suburb Minto

Additional info Centennial Stadium

Tree species Melaleuca styphelioides




Buildings (%) 0.28

Blue space (%) 0.00




73


Latitude -34.0292

Longitude 150.8614

Street Eagle View Road

Suburb Minto

Additional info

Tree species Corymbia torreliana




Buildings (%) 17.04

Blue space (%) 0.00





Latitude -34.0288

Longitude 150.8820

Street Hansens Road

Suburb Minto

Additional info End of Hansens Road, adjacent to fire trail gate





Buildings (%) 0.00

Blue space (%) 0.00




75


Latitude -34.0287

Longitude 150.9055


Suburb Holsworthy

Additional info Base and right of Long Point

Tree species Eucalyptus sclerophylla




Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.0484

Longitude 150.8835

Street Keith Longhurt Reserve

Suburb Kentlyn

Additional info On your left after the last property on Georges River Road

Tree species Acacia spp.




Buildings (%) 1.70

Blue space (%) 0.00




77


Latitude -34.0442

Longitude 150.8653

Street Newman Road

Suburb Kentlyn

Additional info Turnaround circle at end of Newman Road





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.0469

Longitude 150.8416

Street Closest access from Northam Road

Suburb Leumeah Park

Additional info Leumeah Park Tree along back of houses

Tree species Grevillea robusta




Buildings (%) 21.93

Blue space (%) 0.00




79


Latitude -34.0471

Longitude 150.8244

Street Lorraine Cibilic Reserve

Suburb Woodbine

Additional info access from North Steyne Road; tree is right next to North Steyne Road





Buildings (%) 1.41

Blue space (%) 0.00





Latitude -34.0454

Longitude 150.8085

Street Gidley Crescent

Suburb Claymore

Additional info





Buildings (%) 15.05

Blue space (%) 0.00




81


Latitude -34.047

Longitude 150.7773

Street Mt Annan Christian College

Suburb Currans Hill

Additional info





Buildings (%) 0.62

Blue space (%) 0.00





Latitude -34.046746


Street Corner of Currans Hill Drive and Hartley Road

Suburb Currans Hill

Additional info





Buildings (%) 0.62

Blue space (%) 0.00




83


Latitude -34.065235


Street Water tower in William Howe Regional Park

Suburb Mt Annan

Additional info





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.0646

Longitude 150.7641

Street Corner of Stenhouse Drive and Solander Place

Suburb Mt Annan

Additional info





Buildings (%) 25.80

Blue space (%) 0.26




85


Latitude -34.0647

Longitude 150.8228

Street Corner of Moore and Warby Streets

Suburb Campbelltown

Additional info Tree in medium strip

Tree species Eucalyptus pllularis




Buildings (%) 25.12

Blue space (%) 0.00





Latitude -34.0631

Longitude 150.8242

Street Tyler Street apartments

Suburb Campbelltown

Additional info

Tree species Melaleuca quinquinervia




Buildings (%) 43.58

Blue space (%) 0.00




87


Latitude -34.0656

Longitude 150.8421

Street Flinders Street

Suburb Ruse

Additional info close to corner with Giles

Tree species Callistemon cultivari




Buildings (%) 28.76

Blue space (%) 2.23





Latitude -34.0661

Longitude 150.8621

Street Hamilton Road

Suburb Kentlyn

Additional info near corner with Georges River Road

Tree species Eucalyptus oblonga




Buildings (%) 4.31

Blue space (%) 0.00




89


Latitude -34.0802

Longitude 150.8612

Street

Suburb Kentlyn

Additional info Follow 1.4 to a T section, take the right. The pink ribbon indicates to look straight in that area





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.0825

Longitude 150.8662


Suburb Holsworthy

Additional info Parallel from Kentlyn Public School





Buildings (%) 0.00

Blue space (%) 0.00




91


Latitude -34.0824

Longitude 150.8422

Street

Suburb

Additional info All of this fire trail route is passable for a vehicle

Tree species Angophora floribunda




Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.0836

Longitude 150.8233

Street Corner of Docharty Street and Karingal Place

Suburb Bradbury

Additional info

Tree species Melaleuca bracteata




Buildings (%) 18.32

Blue space (%) 0.00




93


Latitude -34.0839

Longitude 150.8055

Street 85 Georgiana Crescent

Suburb Ambervale

Additional info On the corner of Throsby way





Buildings (%) 21.70

Blue space (%) 0.63





Latitude -34.0843

Longitude 150.7888

Street Glen Alpine Golf Course

Suburb Glen Alpine

Additional info near hole 15

Tree species Fraxinus angustifolia




Buildings (%) 1.11

Blue space (%) 0.37




95


Latitude -34.0832

Longitude 150.7668

Street Mt Annan Botanical Gardens

Suburb Mt Annan

Additional info Tree near Wattle Gardens





Buildings (%) 0.90

Blue space (%) 0.00





Latitude -34.0697

Longitude 150.7669

Street Mt Annan Botanical Gardens

Suburb Mt Annan

Additional info Tree behind playground





Buildings (%) 1.38

Blue space (%) 0.00




97


Latitude -34.0879

Longitude 150.7363

Street Elizabeth Macarthur, north

Suburb Camden Park

Additional info

Tree species Eucalyptus tereticornis/amplifolia




Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.1015

Longitude 150.7321

Street Elizabeth Macarthur, middle

Suburb Camden Park

Additional info

Tree species Eucalyptus baueriana




Buildings (%) 0.00

Blue space (%) 0.00




99


Latitude -34.1023

Longitude 150.7453

Street Station Road

Suburb Menangle Park

Additional info Tree on Menangle Park station side across from station road





Buildings (%) 2.20

Blue space (%) 0.00





Latitude -34.1022

Longitude 150.7666

Street Broughton Anglican School


Additional info





Buildings (%) 3.49

Blue space (%) 0.00




101


Latitude -34.1002

Longitude 150.7864

Street 37 Demetrius Reserve

Suburb Rosemeadow

Additional info sensor found in the ground





Buildings (%) 1.24

Blue space (%) 0.00





Latitude -34.1023

Longitude 150.8040

Street 161 Kellermen Drive

Suburb St Helens Park

Additional info

Tree species Eucalyptus baueriana




Buildings (%) 9.46

Blue space (%) 0.00




103


Latitude -34.1016

Longitude 150.8224

Street 2 Lynwood Road

Suburb Lynwood

Additional info Lynwood Park





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.10305

Longitude 150.84329

Street Jeans Fire Trail

Suburb Holsworthy

Additional info Old number 65, I sector





Buildings (%) 0.00

Blue space (%) 0.00




105


Latitude -34.11627

Longitude 150.8390

Street Aberfoyle Fire Trail

Suburb Holsworthy

Additional info Old number 66, I sector

Tree species Corymbia gummifera




Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.12189

Longitude 150.83273

Street Aberfoyle Road

Suburb Holsworthy - Wedderburn

Additional info Old number 67, just inside the fire trail gate





Buildings (%) 0.00

Blue space (%) 0.00




107


Latitude -34.1249

Longitude 150.7444

Street Menangle Station

Suburb Menangle

Additional info





Buildings (%) 4.42

Blue space (%) 0.00





Latitude -34.1376

Longitude 150.7433

Street Ellel Ministeries

Suburb Menangle

Additional info





Buildings (%) 0.00

Blue space (%) 0.00




109


Latitude -34.139

Longitude 150.7819

Street

Suburb Gilead

Additional info Beuleah off the gravel road





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.1398

Longitude 150.8063

Street Bellbird Road

Suburb Wedderburn

Additional info

Tree species Eucalyptus eugenoides




Buildings (%) 0.44

Blue space (%) 0.00




111


Latitude -34.1362

Longitude 150.8240

Street 12 Hodgson Close

Suburb Wedderburn

Additional info

Tree species Corymbia gummifera




Buildings (%) 1.65

Blue space (%) 0.00





Latitude -34.1634067


Street

Suburb Wedderburn

Additional info First pool after Jingga

Tree species Ceratopetalum apetalum




Buildings (%) 0.00

Blue space (%) 14.88




113


Latitude -34.156

Longitude 150.8239

Street 70 Victoria Road

Suburb Wedderburn

Additional info





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.1579

Longitude 150.8064

Street Lot 10 Exley Road

Suburb Wedderburn

Additional info





Buildings (%) 0.00

Blue space (%) 0.00




115


Latitude -34.1579

Longitude 150.7451

Street 530 Moreton Park Road

Suburb Douglas Park - Menangle

Additional info





Buildings (%) 0.00

Blue space (%) 0.36





Latitude -34.175

Longitude 150.7405

Street Morton Park Road

Suburb Douglas Park - Menangle

Additional info Across from 300 Morton Park Road





Buildings (%) 0.00

Blue space (%) 0.00




117


Latitude -34.1806

Longitude 150.7842

Street Brian Road

Suburb Appin

Additional info





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.1746

Longitude 150.8091

Street Minerva Road

Suburb Wedderburn

Additional info Tree on the left of the airport gate, at the end of





Buildings (%) 0.00

Blue space (%) 0.00




119


Latitude -34.17448124


Street 10B Fire Trail

Suburb Appin

Additional info Dharawal NP, stag tree (dead)

Tree species




Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.17728262


Street 10D Fire Trail

Suburb Appin

Additional info Dharawal NP, 100m from landslide

Tree species Banksia cuninghamii




Buildings (%) 0.00

Blue space (%) 0.00




121


Latitude -34.1056

Longitude 150.795

Street Copperfield Drive

Suburb Rosemeadow

Additional info Tree inside Rosemeadow Primary school facing Copperfield Drive





Buildings (%) 5.01

Blue space (%) 0.00





Latitude -34.0744

Longitude 150.8095

Street Apex Park

Suburb Bradbury

Additional info

Tree species Eucalyptus amplifolia




Buildings (%) 0.27

Blue space (%) 0.00




123


Latitude -34.0691

Longitude 150.8119

Street Queen Street

Suburb Campbelltown

Additional info Pedestrian walkway outside the Mall near historic area





Buildings (%) 41.71

Blue space (%) 0.00





Latitude -34.0662

Longitude 150.8132

Street Milgate Arcade car park

Suburb Campbelltown

Additional info

Tree species Eucalytpus sideroxylon




Buildings (%) 0.00

Blue space (%) 0.00




125


Latitude -34.0662

Longitude 150.7841

Street Ehrlich Street Corner of King Street

Suburb Macarthur Heights


Tree species




Buildings (%) 19.64

Blue space (%) 0.00





Latitude -34.0709

Longitude 150.7822

Street Near Macarthur Heights Playground



Tree species




Buildings (%) 18.54

Blue space (%) 0.00




127


Latitude -34.072169


Street Galileo Street



Tree species




Buildings (%) 44.12

Blue space (%) 0.00





Latitude -34.0763

Longitude 150.781

Street University Drive Corner of Cosgrove Street



Tree species




Buildings (%) 6.22

Blue space (%) 0.00




129


Latitude -34.0735

Longitude 150.7842

Street Macarthur Heights


Additional info

Tree species Eucalyptus spp.




Buildings (%) 0.00

Blue space (%) 4.15





Latitude -34.1092

Longitude 150.7482

Street Menangle Park Greenfield - off Menangle Park Rd


Additional info

Tree species Eucalyptus mollucana (unsure)




Buildings (%) 0.00

Blue space (%) 0.00




131


Latitude -34.1176

Longitude 150.786

Street Bottom left corner of Noorumba Reserve, Rosemeadow

Suburb Gilead

Additional info





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.078489

Longitude 150.72845

Street Camden/Nepean River

Suburb Spring Farm

Additional info

Tree species Large leafed privet




Buildings (%) 0.00

Blue space (%) 14.52




133


Latitude -34.1181

Longitude 150.7288

Street Elizabeth Macarthur - south, first point

Suburb Menangle

Additional info





Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.1571

Longitude 150.7894

Street Appin Road

Suburb Gilead

Additional info Horse paddock





Buildings (%) 0.00

Blue space (%) 0.00




135


Latitude -34.1229

Longitude 150.8001

Street Wedderburn River

Suburb Gilead

Additional info

Tree species Acacia longifolia (subsp. longifolia)




Buildings (%) 0.00

Blue space (%) 0.00





Latitude -34.1758

Longitude 150.7741

Street Inghams Road

Suburb Appin

Additional info





Buildings (%) 0.00

Blue space (%) 0.00




137


Latitude -34.046389



Suburb Holsworthy

Additional info Parallel river location, the Basin Kentlyn

Tree species




Buildings (%) 0.00

Blue space (%) 0.00






139

WESTERNSYDNEY.EDU.AU

CONTACT USWestern Sydney University,

Locked Bag 1797, Penrith, NSW 2751

the impact of surface cover & tree canopy on air temperature ...

Documents