School of the Environment, Flinders University, Adelaide, South Australia, Australia E-mail: [email protected]
National Centre for Groundwater Research and Training, Adelaide, South Australia, Australia
School of Chemistry and Physics, Adelaide University, Adelaide, South Australia, Australia
Bureau of Meteorology , Melbourne, Victoria, Australia
Abstract
The urban heat island (UHI) effect significantly raises energy requirements for cooling in cities. The resulting additional use of fossil fuels leads to an increase in greenhouse gas (GHG) emissions. Any means of reducing the strength of the UHI has important ramifications for climate and resources. One approach to achieve this is the establishment of urban green spaces, where transpiration by vegetation consumes some of the additional energy that is produced in the UHI.
Due to its unique, carefully planned layout, the City of Adelaide provides a unique setting for studying the effects of nature reserves on the UHI. The Central Business District (CBD) with densely packed commercial buildings is separated from suburbia by an approximately 500 m wide strip of parks and sporting grounds. Uninterrupted time series data was collected over a period of two years, using a dense network of temperature sensors as well as ten comprehensively instrumented stations. Augmented by vehicle traverses, this record was used to establish the spatial and temporal structure of the near-surface thermal climate.
Results show that the intensity of the nocturnal UHI on an individual night in summer was about 3oC with a maximum intensity of 8oC in the early morning, being most pronounced during clear and calm conditions. Open rural landscapes and parkland cooled faster and more than the urban area. The UHI intensity generally showed a minimum early afternoon and a maximum about six hours after sunset. At the parkland sites, diurnal trends were clearer and stronger than in the CBD.
Keywords: temperature network, traverses, green spaces, parkland belt
Introduction
The increasing worldwide urbanisation requires urban planners to account for the the anthropogenic urban heat island (UHI) effect, Coutts et al. (2007). A major portion of the world’s greenhouse gas (GHG) emissions is produced by cities and urban areas. The energy used for cooling and heating of buildings accounts for most of these GHG emissions (Grimmond, 2007). The mitigation of the urban heat island effect is achievable through an extension of green areas, open spaces and especially parkland (Yokohari et al., 1997).
The city of Adelaide provides a unique setting for studying the effect of parkland on the UHI. The Central Business District (CBD), roughly 2 km by 2 km square, consists of a region of relatively dense high rise buildings surrounded by an approximately 500 m wide strip of parkland. The parkland is generally covered with short grass and sparse trees. Vinodkumar et al. (2012) suggest that these parkland are probably responsible for mitigating the heat island.
In 2010 Flinders University researchers from the School of the Environment supported by a range of South Australian governmental departments and other local funding agencies established an observational network to capture the spatial and temporal status of the urban temperature field. The setup of the network and a higher resolution traverse project is detailed in chapter 2, while some results from the now more than two years of data are displayed in chapter 3. Concluding remarks are given in chapter 4.
Methodology
The Temperature Network
Between July 2010 and November 2010 a total of 39 locations were equipped with miniature encapsulated temperature sensor/logger units, the Thermochron iButton series DS1921H-F5, DS1921Z-F5 and the DS1922L- F5, manufactured by Maxim Integrated Products (Dallas, Texas). The sensors were mounted in custom-made radiation shields at a height of 4m above ground in Adelaide’s Central Business District, the surrounding parkland and the suburbs (blue dots, Fig.1). In addition data from ten stations from the Bureau of Meteorology (BoM) and the Environmental Protection Agency (EPA) could be analysed.

The Traverses
In the southern hemisphere summer and autumn 2010/11 a total of four traverses were performed. For the traverses temperatures at heights of 30cm and 1.8m above ground, surface temperature, net-long-wave radiation, sky temperature and temperature and humidity at 1.6m were measured. The sensors were mounted on a pole attached to the roof rack of a car. The tracks were predominantly oriented south-north and east- west through the city centre, the parkland (green dashed lines, Fig.1) and out into the suburbs.
Synoptic Situation
During the traverses a high pressure cell was located east of Tasmania moving away from the continent with a cold front approaching Adelaide from the south-west. Over Adelaide this resulted in weak easterly winds during the first part of the measurements and hence conditions were favourable for an urban heat island development.
The EPA monitoring weather station Netley, located close to the Adelaide airport, but on the city site in the suburb of Netley, exhibits a change at approximately 11:30pm, the wind speed dropped and the winds turned westerly. A modification effect on the development of the UHI was not found.

Data Analysis
The traverse data as well as the iButton data were adjusted for known calibration errors. Before the iButtons were deployed in the field and after recovering the buttons a correlation study was performed against a well calibrated sensor in order to establish the differences between each sensor (measurements have to be very accurate to account for small changes found in temperature in the urban environment) and to verify the stability of the measurements over time. Temperature, humidity and radiation traverse data were adjusted in time to GPS time.
The traverse instrumentation was mounted in front of the car, in order to avoid a heat effect of the car. To reduce the effect of cars driving in front, low car speeds were rejected from the data set. Travel times and locations were evaluated to find coinciding data from the fixed network.
Results and Discussion
Diurnal Cycle in CBD and Parkland
On the 19th/20th January 2011, the temperature difference between the CBD and the surrounding parkland (Fig.2) is greatest in the early morning between 4:30 am and 6 am. The difference between individual stations e.g. Currie St, which is surrounded by the highest buildings in the Adelaide CBD and the Southern Parkland (Fig.2) reached up to 8°C during this particular night.
Fig.3 displays the spatial distribution of the temperature, acquired from the iButton network. Here the highest temperatures were found where the lowest sky-view factor exists. The green areas, parkland and Victoria Square (a green space in the centre of the CBD) exhibit lower temperatures (Fig.3).

The traverse of the 19th/20th January 2011
Over a six hour period during the night of the 19th to the 20th of January 2011 a number of north-south and east-west traverses of the CBD were collected. Preliminary analysis shows a good agreement between the traverse data and coinciding network data. Fig. 4 exhibits a short, 4 minute long, track from North Adelaide towards Victoria Square, crossing the Torrens River.

The temperature at about 12:00 am increases towards the CBD centre where the highest buildings are, decreases towards the Torrens and exhibits an increase into North Adelaide. This is also visible at 5:30 am in Fig.3 with higher temperatures in the CBD, a local minimum along the Torrens River and a local maximum again in North Adelaide. Fig.5 does show the maximum in the CBD but does not reveal the second maximum in North Adelaide. Although the lower temperatures further north were represented in the traverse data.

Conclusion
The temperature network in Adelaide’s CBD can clearly exhibit the nocturnal urban heat island. The area in the city with highest buildings and hence largest heat storage capacity does reveal higher temperatures during low wind and clear sky conditions.
Traverses through the parkland and city area display an even larger variation in temperature than the fixed network can resolve. Local minima, e.g. the Torrens River, are generally not represented well in the fixed network but were analysed from the traverse data.
From Figures 3 and 5 it is clear that even small open spaces like Victoria Square mitigate the heat island. This square is located in the centre of the CBD, but generally displays a local minimum in the temperature data.
The iButton temperature network which is running for approximately two years. A nocturnal urban heat island develops in most of the weak wind and clear sky conditions with winter and summer differences between city and parkland off similar magnitude [4]. While daytime temperatures in summer and winter in CBD and parkland reveal significant climatic distinctions.
Acknowledgements
Funding was provided by South Australian departments and some local governmental organisations, namely DENR, ACC, DPC, DPLG, ETSA. The EPA and the BoM provided data from their local station network and the University of Adelaide and Flinders University provided support and funding. Graeme Hopkins, Craig Simmons and Peter Schwerdtfeger contributed at the early stage of the project . Chris Kent, Chuanyu Zhu and Robert Andrew assisted with the field data collection.
References
- 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. J. Appl. Meteor. Climatol., 46, 477– 493.
- Grimmond S (2007) Urbanization and global environmental change: Local effects of urban warming. The Geographical Journal, 173, 83−88 Maxim Integrated Products, Inc. 160 Rio Robles San Jose, CA 95134 USA http://www.maxim-ic.com/products/ibutton/ [30 June 2012]
- Vinodkumar H, Guan C T, Simmons J M, Bennett C M, Ewenz C, Kent (2012, submitted). Influence of parks on urban environment: Observational and numerical study over Adelaide, Australia.
- Yokohari M, Brown R D, Kato Y, Yamamoto S (1997) The cooling effect of paddy fields on summertime air temperature in residential Tokyo, Japan. Landscape Urban Plan, 53, 17 – 27.