Urban heat islands

The urban environment modifies microclimate in numerous ways, the net result being termed the urban heat island effect. In general, urban climates are warmer and less windy than in rural areas. However, the modification to urban climates is highly variable, and depends on the particular topography, regional wind speeds, urban morphology, and many other factors.

The majority of research into urban heat islands has been on climates with hot summers. To overcome the lack of data for temperate climates, researchers at the BRE and Brunel University have studied the summer urban heat island of London.

The alteration to temperature is particularly important to building users, as it directly affects fuel use in buildings. The energy used for cooling in the summer will rise (including night cooling) while energy used for heating in the winter will (to a less extent) decrease. However, the degree to which increased cooling demand is offset by lower heating needs varies considerably.

The form and fabric of an urban area differ substantially from rural zones, and this alters the way in which heat flows into and out of the area. Solar gain is more efficiently captured in a city and, together with heat from anthropogenic sources, is better stored. Likewise the way in which heat may be dispersed through re-radiation and air cooling is less effective than in the countryside.

Heat island effects

The effect of these urban rural differences can be seen if temperature is measured along a transect from the countryside into a city. Overleaf, figure 1 shows (in idealised form with clear skies and light winds) the spatial variation in temperature for a large city¹.

On clear, calm evenings after sunset, the temperature rises sharply at the outskirts of the urban area, broadly plateaus across suburbia, and rises gently to the peak of the urban core.

The position of the peak temperature of a heat island is not static, but is typically at the urban core. However, on individual days it will shift downwind if the wind is blowing. Moreover, certain cities have a reduction in heat island intensity at the geographical centre because of some temperature lowering feature such as a lake or a park1.

The difference between urban and rural temperatures is highly dependent on the time of day. Figure 2 shows this diurnal variation in urban-rural temperature difference, again in ideal conditions.

Urban environments are usually warmer before sunrise than rural ones, because they have cooled down overnight at a slower rate. This is primarily because clusters of buildings increase an area's heat capacity and reduce its radiative cooling efficiency. As the sun rises, solar energy evaporates dew in rural areas but starts heating up urban fabric immediately. When the dew has evaporated, the rural surfaces start to warm up more rapidly than the urban ones, due to their lower heat capacity.

As the day progresses the urban and rural temperatures converge and may cross over in summer months when anthropogenic heat release in cities is small. At, or before sunset, the rural surfaces rapidly cool, the rate reducing as dew is formed and latent heat released. Cities tend to cool more slowly and the urban to rural temperature difference starts to rise. The time of maximum heat island intensity varies, but is usually a few hours after sunset.

This scenario, however, is not universal. Because of the multiplicity of factors that contribute to urban temperature, it can be the case that particular cities have unique characteristics.

At wind speeds above about 5 m/s, temperatures across urban environments become more homogeneous and heat islands may not be discernible.

Detailed investigations of the urban climate were carried out in London in the 1960s. This found that London was usually warmer than its rural surroundings on summer days by up to 4°C, but for one third of the time it was cooler².

This study is looking at the temperatures in London, the urban heat island and its spatial variation, and impact on air conditioning. As part of the research, an experiment has been set up to characterise the London urban heat island.

Approximately 70 measurement stations have been installed, arranged on eight transects through London with the central point in Bloomsbury. Temperatures are recorded simultaneously every hour. Data from the summer of 1999 have been collected and measurements are continuing through the summer of 2000.

Results from the London study

Early results show that the intensity of London's heat island varies considerably in time and space. The core temperature has been recorded on occasion as being from -7°C to +7°C warmer than a rural site west of London. However, typical values are less extreme and the average of all the hourly data, irrespective of time of day, between May 31 and August 31 1999, showed that London was warmer than the reference rural site by 1·8°C.

London is usually a classic heat island in that it is much warmer than rural areas at night, but less so during the day. Examples of the temperature distribution across London are shown in figures 3 and 4. Both these figures have temperature contours spaced at 0·5°C intervals, but the colours of the temperature bands are different. During this day the wind speed, as measured at a north western site, was low (0-2 m/s). In the early morning, a 5-6°C difference can be seen between the warmest part of London's core temperature and the rural areas.

Later the same day, at 16.00 h, the rural areas have heated up faster than the urban areas, and the temperature across the area is more homogeneous. Nevertheless, there are temperature differences of about 3°C, and 5°C in places, particularly in the north.

Impact on cooling energy use

In summer, the urban heat island has an impact on both air-conditioned and passively cooled buildings, such that:

increased external temperature results in an increased energy demand for cooling

night-time cooling potential is reduced

exceptionally hot periods can lead to a dramatic rise in the demand for air conditioning, increasing the peak electricity demand

air conditioning systems are less efficient when operating under large temperature differences

the profile of temperature and wind in an urban environment is modified and this affects the efficacy of passive cooling designs

Reducing the impact

There are two recognised ways of reducing the intensity of an urban heat island. First, by increasing the albedo of the urban fabric and second, by increasing the amount of vegetation, particularly trees.

These measures respectively reduce the absorption of solar radiation and increase evaporative cooling. The absorptivity of urban areas is greater than in rural areas. Surfaces are generally darker, and the urban gorges between buildings offer multiple opportunities for absorption at each reflection of incoming radiation. Hence the urban reflectivity to solar radiation – the albedo – can be increased by using lighter coloured roofing material.

Modelling studies have indicated that if the average albedo of the urban areas in the Los Angeles Basin were increased from 0·13 to 0·26, central Los Angeles would become 2-4°C cooler at midday in the summer. Peak power use would fall by 0·6-1·2 GW³.

Trees offer shade from direct radiation, which can improve comfort to pedestrians and reduce the radiation absorbed on the ground. The evapotranspiration of all vegetation directly cools the air.

A combined measurement and modelling study (in Tokyo) has shown that cool air advected from a park to a commercial area had a potential of reducing air-conditioning by almost 15% between 13.00 h – 14.00 h⁴.

The effects of the urban heat island on energy demand in European cities, and possible solutions, are included in a recent publication⁵.

The study will continue, using thermal and airflow simulations on a representative building positioned in a number of locations to investigate various scenarios. Using the field measurements and computer modelling results, guidelines will be formulated on ways of moderating the effect of the heat island on cooling energy demand.

The findings of the work will be included in two of the major outputs: a guidance document and a design tool for the construction industry to enable them to move towards reducing the impact of the urban heat island on peak cooling energy demand in buildings. This is scheduled to be produced in draft form by February 2001.

John Palmer and Paul Littlefair are with the Environmental Engineering Centre at the BRE. Richard Watkins and Maria Kolokotroni are with Brunel University.

This article is based on the DETR funded project 'Site layout planning to improve solar access, passive cooling and microclimate'.