Cities as agents of global change - Wiley Online Library

INTERNATIONAL JOURNAL OF CLIMATOLOGY Int. J. Climatol. 27: 1849–1857 (2007) Published online 22 August 2007 in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/joc.1604

Cities as agents of global change Gerald Mills* School of Geography, Planning & Environmental Policy, Newman Building, UCD, Dublin 4, Ireland

Abstract: Cities, as places where human activities are concentrated, are frequently cited as the chief causes of, and solutions to, anthropogenic global change. In this article, I review the climatology literature that examines the relationship between cities and atmospheric changes at all scales. Despite the volume of literature on this theme, there is little overall coherence. In part, this is a result of the varying operational definitions of the city and the difficulty in obtaining pertinent information. Rather than attempt to provide a comprehensive review of the literature that focuses on cities and global change, this article categorises published research on the relationship between urban areas and climate changes at all scales into common themes. Copyright  2007 Royal Meteorological Society KEY WORDS

urban climate; global change; cities

Received 26 September 2006; Revised 19 June 2007; Accepted 22 June 2007

INTRODUCTION We are at an important juncture of human relationship with the earth. Approximately half the world’s population (3.2 billion) now lives in urban areas, and this proportion will grow substantially in the near future. By 2030, it is expected that 5 billion will live in urban areas and account for 60% of the global population. The greater proportion of this growth is expected to occur in the less developed world, particularly in Asia, where 71% of the global rural population currently resides. Although some of this growth is predicted to occur in very large cities, most is expected to occur in smaller cities with populations less than one million. In 2005, cities with populations under 500 000 accounted for 51% of urban dwellers (UN, 2006). While we know something of the larger cities and the problems they experience, little is known of the smaller cities where most will reside. ‘Urbanization’ is used to refer both to the movement of people into cities and to the transformation of ‘natural’ into urban land-cover. An urban land-cover consists of closely spaced buildings, impervious surfaces and managed outdoor spaces. Currently, just 2–3% of the earth’s ice-free land area corresponds to this definition – yet half of humanity resides in these areas, and it is here that human activity is concentrated. Currently, cities are the foci for the planetary flows of energy and materials, which are used to construct the physical city and sustain its functions (Decker et al., 2000). As a consequence, cities are directly and indirectly responsible for global changes in the atmosphere, hydrosphere, geosphere and * Correspondence to: Gerald Mills, School of Geography, Planning & Environmental Policy, Newman Building, UCD, Dublin 4, Ireland. E-mail: [email protected] Copyright  2007 Royal Meteorological Society

biosphere. The direct global impact of urban characteristics is best illustrated with two examples. The first example illustrates that the anthropogenic emissions of carbon dioxide are highly concentrated in and near urban centres. Figure 1 illustrates global emissions of carbon dioxide due to fossil fuel use calculated for 1 × 1° latitude–longitude grid (Andres et al., 1997). One is struck by the concentration of these emissions in the Northern hemisphere, and at readily identifiable centres. The geography of emissions corresponds closely to the locations of wealthy cities. For example, note the distribution of carbon injections into the atmosphere in western Europe, the tall spikes in Japan and the lower peaks in eastern China. By contrast, much of Africa and India is characterized by the absence of these carbon peaks, despite the enormity of some cities (e.g. New Delhi and Lagos) located in these regions. The global economic disparity is revealed here as a disparity in urban-sourced emissions. The second example shows that, despite their limited physical imprint, cities can have a direct global impact. While urban settings may provide extraordinarily diverse environments, the urban environment is designed primarily to service human needs exclusively. Thus, cities are biologically homogenizing environments to which only certain species can adapt – in fact, ‘urbanization is one of the leading causes of species extinction’ (McKinney, 2006). While individual cities may appear to be oases of biodiversity, much of this biodiversity is exotic in nature and exists at the expense of local species. Moreover, because cities produce similar urban environments, this urban diversity is common to many cities such that overall, cities serve to diminish global biodiversity.

1850

G. MILLS

Figure 1. Estimated carbon dioxide emissions in 1995 in 000 s of metric tons (t) of carbon per 1° × 1° latitude/longitude grid cell. These values are the summed emissions from fossil-fuel burning, hydraulic cement production and gas flaring. The highest value recorded is approximately 70 000 Mt. (See Andres et al., 1997).

In this paper, I review current published research that links cities (and urbanization) to atmospheric changes at all scales. Initially, I will examine the definition of the city and stress the distinction between two related aspects of the urban effect: 1. Urban stock effect, which is created by the physical presence of the city and, 2. Urban flux effect, which is created by the activities associated with cities. I will use this distinction throughout this paper to identify themes that have emerged in the recent literature on urban–global climate links and explore gaps that remain. Over the last four decades, urban environmental studies have evolved from considerations of processes and impacts associated with individual urban areas (e.g. Wolman, 1965) towards more comprehensive assessments of the totality of the urban effect on global flows of materials and energy (e.g. Decker et al., 2000). This evolution reflects both improved scientific capabilities for observing and detecting anthropogenic effects and the increasing dominance of cities as focussed areas of anthropogenic activities. This review stresses the need to define what is meant by urban terms (such as ‘city’ and ‘urbanization’) as this affects the structure of the study, the results acquired, the policy implications that arise and our ability to communicate.

DEFINING AND MEASURING THE CITY Much of the climate literature assumes that terms like ‘city’ and ‘urban’ are universal. However, establishing Copyright  2007 Royal Meteorological Society

a clear definition is critical to how one measures and manages the urban effect. Leaving aside the question of how to demarcate the boundary of a city, we may first of all distinguish between a definition based on the urbanized land-cover and one based on the urban system. The former identifies the extent of contiguous urban landcover, which normally consists of impermeable materials and closely spaced buildings. It also includes green areas (isolated trees, managed parks and derelict land) embedded within the fabric. In many western cities, the intensity of urban cover decreases from the centre, and the edge becomes difficult to identify as tracts of non-urban cover (e.g. green parks, forested areas and agricultural land) are interwoven with the urban cover. This built-up area is three-dimensional in form and often displays a consistent topography with the tallest structures to be found in the city centre and smaller and widely spaced buildings toward the urban edge. In the poorer cities of the world, the city may be surrounded by informal settlements that may have little infrastructure (e.g. paved roads) and few urban services (e.g. waste collection). Examining the city as an urban system will describe a different place. A systemic approach would link the built-up area to its economic hinterland, the area from which it routinely draws its working population and with which it functions as an economic entity. This city region may consist of a variety of settlements of varying size and extent that are connected via transport and information corridors along which people, goods and information flow. These ‘cities’ are themselves components in wider national and international networks of flows. Thus, for example, a citizen living in a ‘rural’ setting may be as integrated into the urban Int. J. Climatol. 27: 1849–1857 (2007) DOI: 10.1002/joc

CITIES AS AGENTS OF GLOBAL CHANGE

economy as her urban counterpart. For example, consider the implications inherent in Figure 2, which shows the distribution of commuters in Ireland and the urban landcover. The impervious urban surfaces appear as the darkest areas on this map. Recent economic growth in Ireland has resulted in rapid population growth. However, while employment has remained concentrated in urban areas, much of the working population resides outside the physical limits of the urbanized area. This effect is demonstrated by the ‘halo’ present around each major urban centre as citizens seek inexpensive homes and a more ‘rural’ lifestyle. Despite their location, this population clearly ‘belongs’ to the city at the centre. Practically, it is difficult to gather information using either definition. The most commonly employed methods for gathering data are either based on national censuses or on information obtained from remote sensing. The advantages of the former are that they provide authoritative data on administrative areas, which may be used to establish trends in population, economy and the built environment. This information can often be supplemented by other data on lifestyle, water use, etc. to provide a comprehensive picture of the urban citizen. However, administrative units rarely match the physical dimensions of the city and there is no international consistency. In addition,

1851

much basic information is unavailable for many parts of the world. Remote sensing, using satellite technology, offers the potential to provide a standard approach to the routine gathering of limited types of urban information – such as urbanized landscape – on a global basis (Miller and Small, 2003). However, it can provide little information on the three-dimensional character of the city or the nature of its occupation. There are other forms of remote sensing (e.g. LIDAR) that can fill in some of these gaps, however, these are not widely available. It remains the case that our information on cities generally is inconsistent and incomplete and that no single source of information is adequate. These difficulties are well illustrated in a study on the role of urban areas in the natural global carbon cycle (Svirejeva-Hopkins et al., 2004). The researchers examined the extent of ‘urbanized territories’ (that is, the suite of land-cover changes associated with urbanization) across different economies. For each, the authors estimated the changes to the natural carbon flux that ensue as soil is covered and vegetation is removed during urban development, often followed by extensive revegetation as cities ‘mature’. With respect to the natural exchanges of carbon, mature cities (such as those in the developed world) are net removers, while rapidly

Figure 2. Distance to work in Ireland, 2002. The black areas represent the impervious urban cover based on the CORINE land-cover dataset. The shaded areas are those places where between one-fifth and one-half of the population travel more than 30 km to work. Copyright  2007 Royal Meteorological Society

Int. J. Climatol. 27: 1849–1857 (2007) DOI: 10.1002/joc

1852

G. MILLS

growing/expanding cities are net producers, of carbon via natural pathways.

regulated by the three-dimensional urban form, which may act to ventilate or recirculate emissions into this layer.

THE URBAN CLIMATE EFFECT

Urban stock effect The physical effect of the city is due to its transformation of the land surface, which alters the surface energy and water budgets most directly. For the most part, the materials employed to construct cities are impermeable, so that rainwater is quickly removed from the surface and little is available for evaporation. The result is a change in the local hydrology wherein river systems experience large fluctuations in discharge as its flow regime becomes more closely coupled with individual rainfall events. In addition, the sealed surface shifts the emphasis in the energy budget toward sensible heat exchange (rather than evaporation), both with the substrate and with the atmosphere. In addition to its material composition, the urban surface is highly convoluted in terms of geometry. A consequence of this is the creation of a distinguishable UCL, comprised of the airspace below building rooftops. In the ‘open’ UCL, a myriad of climates are produced by complex flow patterns that interact with surfaces that have different orientations, materials and exposures. In the ‘closed’ UCL (that is, building interiors), climates are controlled by regulating energy and mass flows across the material envelope that separates them from the outdoors. Thus, within the UCL the three-dimensional form of the city exerts a profound impact on the micro-scale climate. Above the UCL, separated by a transition layer, lies the urban boundary layer (UBL). The distinctive properties of the UBL are fundamentally derived from the underlying dry, warm, rough and polluting ‘urban’ surface. As airflow crosses the upwind edge of the urban area the overlying air responds and grows in depth as the effects of the urban surface below are mixed with ‘nonurban’ air entrained from above. Once formed, the UBL

One might expect that the field of urban climatology, which is focussed on observing and understanding the urban effect on climate has broached the definitional problem alluded to above. However, this effect is typically measured as the difference between observations at self-described ‘urban’ and proximate ‘rural’ stations. The assumption is that the latter represents the consistent, pre-urban, background climate against which change can be measured. However, Lowry (1977) has shown that, in the absence of in situ observations that pre-date urbanization, one may only estimate the urban effect. More recently, Oke (2004) has argued for detailed ‘metadata’ to accompany observations so that one does not pre-judge the nature of the site from the terms ‘urban’ and ‘rural’, each of which describe environments that are both diverse and dynamic (Stewart, 2006). Given the dual definitions of the city (as physical imprint and as system), we can divide the urban climate effect into that associated with its physical presence (stock) and that associated with its activities (flux). This is not to suggest that they are separate or exclusive. For example, a component of the activities is precisely associated with adding to the physical stock. In a rapidly growing urban area this may form a very significant part of the waste energy and materials emitted to the atmosphere. Figure 3 shows cement production for the five largest cement producing nations. In 1994, rapid development (and urbanization) in China required in 423 mt of cement and resulted in 101.4 mt of carbon emissions, or one-third of the global emissions, from this activity (Worrell et al., 2001). Similarly, the concentration of air pollutants in the urban canopy layer (UCL) is strongly

Figure 3. 1970–1995 cement production trends in the five largest cement-producing countries. The scale on the left is for China only. Redrawn from Worrell et al., 2001. Copyright  2007 Royal Meteorological Society


1853


may undergo significant changes as the accumulated mass of pollutants emitted directly into the atmosphere begin to interact to generate secondary pollutants. Many of these pollutants have distinctive radiative effects so that the thermal properties of the UBL and radiative exchanges with the surface beneath are modified. The relationship between the UBL and UCL is complex as exchanges of heat, momentum and materials are regulated by urban form (and associated activities) and local meteorology. For example, heating by the urban surface overnight is often sufficient to maintain a lapse rate in the UBL temperature profile, thus preventing the development of a nocturnal inversion and maintaining an upward flux of heat and materials. In rare circumstances, the relative warmth of the city surface can generate a local wind system as cooler air is drawn from the surrounding non-urban areas into the city. The distinctive features of the UBL are transferred downwind, beyond the city boundaries. The horizontal extent of the urban effect is regulated both by the nature of the urban surface and the ‘natural’ background climate. However, the distinctive climatic parameters of turbulence and warming dissipate quickly, particularly when the background climate encourages mixing. Urban flux effect This is the result of the operation of the city and accounts for the waste heat, materials and gases that are emitted into the urban atmosphere. The generation of this flux varies with the temporal pattern and nature of activity and of the energy that underpins it. The climatic significance of the anthropogenic flux that results depends upon the magnitude of the natural fluxes. There have been few attempts to directly measure the anthropogenic heat flux in urban areas – an assessment for Toulouse, France suggests an annual average value of 40 Wm−2 , with a winter time maximum of 150 Wm−2 (Pigeon et al., 2006). At high-latitudes (or in very dense, lower latitude cities where the anthropogenic flux is highly concentrated) this additional heat flux is significant for the formation of the urban heat island and to the buoyancy of the urban plume, which carries the atmospheric byproducts of the city downwind. The regional and global impact of cities is most directly associated with their individual urban plumes, whose persistence depends on both the residence times of its constituents and the degree of mixing that the plume experiences (e.g. Molina and Molina, 2004). For example, Stohl et al. (2003) managed to identify the emissions from New York City in a pollution plume over Europe. However, its distinctiveness was partly related to the particular weather conditions that allowed this urban plume to develop and then transport it, with little dilution, into the middle and upper troposphere. More generally, air pollution which is advected to an area may contain contributions from many individual urban areas. In Figure 4, the averaged contribution of major urban centres in Asia to sulphur deposition is shown. Although Copyright  2007 Royal Meteorological Society

the highest contributions are to be found in relative proximity to cities, the urban contribution is detectable at a continental scale. Such findings ‘have erased the distinction between air quality and global atmospheric chemistry’ (Molina and Molina, 2004, p. 652). The extent and magnitude of the combined urban effect depends both on the city and the background climate, which can regulate the dilution of the urban plume. Many cities occupy distinctive topographic settings such as river valleys and basins, which may, under certain circumstances, limit dilution, restrict the extent of the plume and increase its effect. Conversely, in turbulent conditions the plume may be widely dispersed and diluted, which will both broaden and lessen the magnitude of the urban impact. Traditionally, the field of urban climatology has concerned itself with the combined effect of both urban flux and stock within the limits of the city. At a larger scale, an increasing body of literature on meso-scale modelling is concerned with the regional air quality effects of urban areas with a view to ascribing individual city ‘responsibility’. At the global level there has been no published research that has attempted to identify the contributions of individual cities to the global concentration of anthropogenic pollutants. However, it is clear that, as observation and modelling methods improve, this is a direction for future research. Global change and cities A perusal of the current literature on global climate change and cities does not convey a comprehensive or coherent body of work. Rather, it is characterized by themes, some of which overlap. Presented below, is my categorization of this work, which I substantiate using examples selected from this literature. City as cause of global climate change The spatial correspondence between the global emissions of CO2 and the major urban economies of Asia, Europe

Figure 4. Percentage contribution to total sulphur deposition due to SO2 emissions from major urban centres in Asia averaged over the 25 years between 1975 and 2000. The shaded areas represent 5, 30 and 50% of the deposition, respectively. Redrawn from Guttikunda et al., 2003. Int. J. Climatol. 27: 1849–1857 (2007) DOI: 10.1002/joc

1854

G. MILLS

and North America illustrated in Figure 1 is unequivocal. However, while the annual fluxes shown in this diagram are estimated from inventories of fuel use, the specifically ‘urban’ atmospheric flux has not been measured until very recently. Figure 5 shows the flux and concentration of CO2 in Mexico City’s UBL and incorporates natural and anthropogenic sources and sinks (Velasco et al., 2005). While the flux varies through the day, the economic pulse of the city is visible in the plateau of high values in the middle of the day, which correspond to emissions from the transport fleet. During the early morning commute when the mixing layer is shallow, the emissions result in a peak of concentration. Throughout the daytime, the emission rate is nearly constant, however, as surface heating generates turbulence and expands the mixing layer, concentrations fall. After sunset, concentrations rise once again although the emission rate declines. Overall, this research suggests an annual CO2 flux of 12.8 kg m−2 , substantially in line with existing measurements made in other developed world cities. However, such observations remain few and far between and expanding this type of

measurement program to other urban environments is an imperative, if focussed urban mitigation strategies are to be employed. Global climate change and city impact Table I presents the state of current knowledge on the impact that global climate changes may have on human settlements based on the IPCC’s Third Assessment Report (McCarthy et al., 2001). What is noticeable is that there are few specifically urban issues that arise. To a considerable extent, this is partly due to the relative coarseness of climate modelling at this scale that cannot explicitly include urbanized areas. Consequently, the main issues that arise from this table reflect on their changed background climates or their location in coastal zones, which place them at risk of flooding due to sealevel rise. Much of the literature in global climate change policy, as it relates to urban areas, focuses on the adaptation to change (e.g. LCCP, 2002) to protect valuable infrastructure and water resources. The exception to this

Figure 5. Average diurnal pattern of CO2 flux (dotted line) and concentration (solid line) measured over an urban neighbourhood in Mexico City during April 2003. Redrawn from Velasco et al., 2005.

Table I. Human settlements impacts, categorized by state of scientific knowledge. This Table is from Chapter 7 in McCarthy et al. (2001).

LOW Level of Agreement HIGH

Established but incomplete

Well established

Increased vulnerability of infrastructure to urban flooding and landslides Tropical cyclones more destructive Fire danger to urban wildland fringe infrastructure increased Sea-level rise increases cost/vulnerability of resource based industry Water supplies more vulnerable

Sea-level increases cost/vulnerability of coastal infrastructure Energy demand sensitive; parts of energy supply vulnerable Local capacity critical to successful adaptation Infrastructure in permafrost regions vulnerable

Speculative Fire damage to key resources increased More hail and windstorm damage

LOW

Copyright  2007 Royal Meteorological Society

Competing explanations Heat waves more serious for human health, resources Non-climate effects more important than climate Heat island effects increase summer energy demand Increased air and water quality problems Amount of Evidence

HIGH



is a concern for how predicted global warming will interact with the urban heat island. Global warming will result in more frequent ‘heat waves’, periods of high temperatures that overload the human body’s thermoregulation system. The warmer air temperatures and lower wind speeds in urban areas will accentuate the associated thermal strain. In addition, one’s capacity to cope with these conditions may be diminished when, as is often the case, high temperatures are combined with poor air quality. But, global warming will also raise wintertime temperatures, thereby reducing the thermal strain caused by low winter-time temperatures. The relationship between mortality and air temperature, particularly as it relates to cities, is still under study and is likely to be very complex. For example, when detailed studies of urban heat waves are undertaken it is clear that, notwithstanding the thermal environment, heat-related deaths are unevenly spread through the population, a feature that may only be explained by invoking socio-demographic, rather than weather, factors (Klinenberg, 2003). Moreover, estimates of heat-related deaths under global warming scenarios do not generally account for adaptation. Keatinge et al. (2000) examined heat-and cold-related mortality in people aged 65–74 living in cold and warm regions across Europe. The results show that the temperature of minimum mortality is higher in those places with warmer summers. Moreover, the impact on mortality of temperatures lower or higher than this value differs for different climates. The results of this study suggest that populations occupying different climates have adjusted to these thermal regimes. Thus, estimates of changes in mortality that are based on changes in the average temperature over a 50-year period and existing temperature-mortality relationships may be less than has been assumed. City as contaminant A consistent theme of global climate change studies is that urban areas ‘contaminate’ meteorological records and obscure the global climate signal. The chief purpose of standard meteorological stations is to be representative of a wide area that remains ostensibly unchanged, so that a consistent set of records from a geographically dispersed network can be employed to examine trends. However, the effect of cities is to alter the climate in their vicinity – in the case of the near-surface air temperature this effect is in one direction, that of warming. Thus, a station that is within the zone of urban influence will store its effect in its measurements. The methodological difficulty in extracting this urban bias has been to establish the extent and the intensity of the urban effect, which varies with changes in weather/climate and with changes in city form and functions (Lowry, 1977). Despite these difficulties, various simple procedures have been developed to estimate the urban effect and remove it from the observational record. For air temperature, a common method has been to substitute urban population size as a ‘surrogate’ for the magnitude of the urban effect. Gallo and Owen Copyright  2007 Royal Meteorological Society

1855

(1999) argue that satellite-based estimates of urbanized areas can provide comparable results and would provide a consistent means of addressing this problem globally. This approach has the merit of employing physically based measures of the urban setting, but is likely to be equally imprecise. The urban temperature effect is the result of a number of processes, many of which (such as street geometry) are not amenable to satellite measurement. An important first step for the removal of urban, among other, effects is a site-specific description of the measurement site, something that is generally unavailable (Oke, 2004). City as model Changnon (1992) suggested that urban climate change could serve as a useful ‘analogue’ for global climate change. In common with global climate change predictions, the urban atmosphere is both warmer and carbon enriched, when compared to the background climate (Figure 5). However, few have employed the urban climate as a realistic current model for the future climate. Figure 6 shows the results of an experiment carried out in the Baltimore region, using four outdoor enclosures in which ragweed was grown. The experiment was designed to assess the impact of climate change on ragweed biology and allergic rhinitis to which a substantial portion of the US population is sensitive. The four sites lie along a transect from rural to urban locations and correspond with consistent changes in ambient air temperature and CO2 concentration. The results show that the pollen season occurs earlier and greater pollen numbers are increased with urbanization. Although rural pollen contained more allergens, the greater urban production meant more allergens were present in the urban setting. The results indicate that climate change on a very small scale (that is, an urban area) ‘can alter plant physiology and reproductive behaviours in ways that are already likely to be affecting human health’ (Ziska et al., 2003, p295). City as solution As urban areas are major contributors to global climate change, there is a large amount of literature on how cities can be managed to moderate and reduce this contribution. For example, there is an increasing literature on the potential of urban vegetation in reducing net urban carbon emissions (e.g. Nowak and Crane, 2002). This research is often placed within a broader ‘sustainable city’ context, which (from an environmental perspective) seeks to make cities more efficient in their use of resources, and reduce the magnitude of flows focussed on cities (e.g. Rees and Wackernagel, 1997; Newman, 1999). As cities often have strong management and planning systems (particularly in the energy-intensive, economically developed world) they can be restructured and reorganized over time (e.g. Rogers, 1997). This field is extraordinarily broad, so I will select one theme here that links urban form and function. Figure 7 shows the travel time budget for a variety of places that represent an income spectrum (Schafer and Int. J. Climatol. 27: 1849–1857 (2007) DOI: 10.1002/joc

1856

G. MILLS

Victor, 1999). The average is 1.1 h per person per day, and the data shows little variation. Although the fraction of travel time devoted to motorized transport rises with income, there is no corresponding increase in the time devoted to travel. If the time budget is fixed, the ability to traverse distances is largely regulated by the transport available. The implication of this graph is that the size of settlement and the distribution of functions within is strongly linked to its transportation network. This is one strand of an argument that proposes ‘compact’ cities – well designed, densely occupied areas of limited extent – are intrinsically more efficient than lowdensity cities, which are costly to service and traverse. Similarly, Lyons et al. (2003), using worldwide city-level data, found that the observed variations in total vehicle kilometers travelled, which is indicative of pollutant emission, could be statistically explained by a single variable, urbanized land area. The implication of this work is that the anthropogenic emissions of cities can be modulated by managing its physical size.

CONCLUSIONS While the study of the climates of urban areas has traditionally been the realm of urban climatologists, it is clear from the literature that the broader urban effect is now a focus of attention for regional and, increasingly global, climate science. This work is largely concerned with the flux of pollutants, which are emitted largely as a consequence of urban-related activities that may or may not occur within the defined urban area. A comprehensive urban climate science will require that our understanding of the urban effect can be communicated to researchers operating at different scales. It is clear from the literature that much research in this area employs (either explicitly or implicitly) definitions for ‘urban’ that differ substantially from one another. Consequently, it is often difficult to see where the science at one scale fits with that at another scale. For example, urban climatologists have long recognized that air temperatures in an urban area are measured within the UCL in

Figure 6. Time course of ragweed pollen production for four sites along an urban transect for 2001 as a function of day of year. Over the period the average CO2 concentration was 389, 399, 501 and 511 ppm for the rural, semi-rural, suburb and urban sites respectively. Compared to the rural site, the average air temperatures at the other sites (in order of ‘urbanization’) were 0.6, 1.1 and 1.9 ° C warmer.

Figure 7. Average per-capita travel time budget from surveys in African villages, 44 cities and 20 nations. (Redrawn from Schafer and Victor, 1999). Copyright  2007 Royal Meteorological Society



which the three-dimensional building geometry exerts a considerable influence on observations. However, global analyses continue to treat the urban landscape as a flat, artificial surface and implicitly place the air temperature observations ‘above’ this surface. An important starting point in any research should be to clearly identify the operational definition for ‘urban’ used in the study and to identify its limitations. It is only recently that urban climtologists have themselves recognized the need to be more explicit in their use of ‘urban’ and to identify different types of urban landscapes that give rise to different urban effects (Oke, 2004). One area where there is considerable scope for interaction across the scales of enquiry is that of climate change mitigation. At the global scale, the anthropogenic CO2 flux arises from relatively small areas: for example in Figure 1, 50% of the total CO2 emission arises from 3% of land area; 10% arises from approximately 1846 km2 . Undoubtedly, if greater geographic detail were obtainable, the bulk of this emission would be shown to arise from within urban administrative boundaries where there are existing planning systems of varying sophistication. It is eminently possible to modify both form (e.g. vegetation and material properties) and activities at this scale to regulate the magnitude of material and energy flows and their attendant wastes (Mills, 2006). There is already a substantial body of urban climate literature that links elements of urban design to climatic outcomes. Unfortunately, much of this research is based on case studies that do not yield general relationships that have widespread applicability. Nevertheless, there is an appreciation in urban climatology for the nature of the urban effect and its measurement that is currently absent in the discussions of the urban effect at the global scale. A desirable research strategy would focus on design at the scale of building groups (components of cities) where urban measures (e.g. building design and vegetative cover) could be linked to discernible climatic effects (e.g. CO2 fluxes). ACKNOWLEDGEMENTS

I am very appreciative of the comments made by three anonymous referees and those of the editor. I hope that the article has been improved by my efforts to incorporate their viewpoints. In addition, I am grateful to Stephen Hannon who assisted in the graphics accompanying the article. REFERENCES Andres RJ, Marland G, Fung I, Matthews E, Brenkert AL. 1997. Geographic Patterns of Carbon Dioxide Emissions from Fossil-Fuel Burning, Hydraulic Cement Production, and Gas Flaring on a one Degree by one Degree Grid Cell Basis: 1950 to 1990. Carbon Dioxide Information Analysis Centre, Oak Ridge National Laboratory: Oak Ridge, Tennessee; Environmental Science Division No. 4646. Changnon SA. 1992. Inadvertent weather modification in urban areas: Lessons for global climate change. Bulletin of the American Meteorological Society 73: 619–627. Decker EH, Elliott S, Smith FA, Blake DR, Rowland FS. 2000. Energy and material flow through the urban ecosystem. Annual Review of Energy and the Environment 25: 685–740. Copyright  2007 Royal Meteorological Society

1857

Gallo KP, Owen TW. 1999. Satellite-based adjustments for the urban heat island temperature bias. Journal of Applied Meteorology 38: 806–813. Guttikunda SK, Carmichael GR, Calori G, Eck C, Woo J. 2003. The contribution of megacities to regional sulphur pollution in Asia. Atmospheric Environment 37: 11–22. Keatinge WR, Donaldson GC, Cordioli E, Marinelli M, Kunst AE, Mackenbach JP, Nayha S, Vuori I. 2000. Heat related mortality in warm and cold regions of Europe: Observational study. British Medical Journal 321: 670–673. Klinenberg E. 2003. Heat Wave: A Social Autopsy of Disaster in Chicago. University of Chicago Press: Chicago, Illinois. LCCP (London Climate Change Partnership). 2002. London’s Warming: The impacts of Climate Change on London. Technical Report, Greater London Authority. Lowry WP. 1977. Empirical estimation of urban effects on climate: A problem analysis. Journal of Applied Meteorology 16: 129–135. Lyons TJ, Kenworthy JR, Moy C, dos Santos F. 2003. An international urban air pollution model for the transportation sector. Transport Research D 8: 159–167. McCarthy JJ, Canziani OF, Leary NA, Dokken DJ, White KS (eds). 2001. Climate Change 2001: Impacts, Adaptation, and Vulnerability. Contribution of Working Group II to the Third Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, UK. McKinney ML. 2006. Urbanization as a major cause of biotic homogenization. Biological Conservation 127: 247–260. Miller RB, Small C. 2003. Cities from space: potential applications of remote sensing in urban environmental research and policy. Environmental Science and Policy 6: 129–137. Mills G. 2006. Progress toward sustainable settlements: A role for urban climatology. Theoretical and Applied Climatology 84: 69–76. Molina MJ, Molina LT. 2004. Megacities and atmospheric pollution. Journal of the Air and Waste Management Association 54: 644–680. Newman PWG. 1999. Sustainability and cities: extending the metabolism model. Landscape and Urban Planning 44: 219–226. Nowak DJ, Crane DE. 2002. Carbon storage and sequestration by urban trees in the USA. Environmental Pollution 116: 381–389. Oke TR. 2004. Initial Guidance to Obtain Representative Meteorological Observations at Urban Sites. World Meteorological Organization, Geneva. Instruments and Observing Methods: Report No. 81. WMO/TD No. 1250. Pigeon G, Durand P, Masson V. 2006. Anthropogenic heat release over Toulouse during the CAPITOUL field program. Proceedings of the 6th International Conference on Urban Climate, Sweden. Rees W, Wackernagel M. 1997. Urban ecological footprints: Why cities cannot be sustainable–and why they are a key to sustainability. Environmental Impact Assessment Review 16: 223–248. Rogers R. 1997. Cities for a Small Planet. Faber and Faber: London. Schafer A, Victor DG. 1999. Global passenger travel: implications for carbon dioxide emissions. Energy 24: 657–679. Stewart I. 2006. Representations of “Urban” and “Rural” Space in UHI Literature. IAUC Newsletter 19, available at www.urban-climate.org, 14–17. Stohl A, Forster C, Eckhardt S, Spichtinger N, Huntrieser H, Heland J, Schlager H, Wilhelm S, Arnold F, Cooper O. 2003. A backward modeling study of intercontinental pollution transport using aircraft measurements. Journal of Geophysical Research 108: No. D12, 4370. Svirejeva-Hopkins A, Schellnhuber HJ, Pomaz VL. 2004. Urbanised territories as a specific component of the global carbon cycle. Ecological Modelling 173: 295–312. UN. 2006. World Urbanization Prospects: The 2005 Revision. United Nations: New York. Velasco E, Pressley S, Allwine E, Westberg H, Lamb B. 2005. Measurements of CO2 fluxes from the Mexico City urban landscape. Atmospheric Environment 39: 7433–7446. Wolman. 1965. The metabolism of cities. Scientific American 213: 179–190. Worrell E, Price L, Martin N, Hendriks C, Meida LO. 2001. Carbon dioxide emissions from the global cement industry. Annual Review of Energy and the Environment 26: 303–329. Ziska LH, Gebhard DE, Frenz DA, Faulkner S, Singer BD, Straka JG. 2003. Cities as harbingers of climate change: Common ragweed, urbanization, and public health. Journal Allergy and Clinical Immunology 111: 290–295. Int. J. Climatol. 27: 1849–1857 (2007) DOI: 10.1002/joc