Urban areas are expected to continue their rapid growth in the twenty-first century. Globally, cities are major sources of greenhouse gases emissions and their high population densities make them potential focal points of vulnerability to global environmental change. Moreover, their reach, in terms of flows of materials and resources, extends far outside their borders. Evidently, it is no longer tenable to consider urban systems to be static artefacts constructed in a stable environment, nor continue to divorce them from the global context that influences many of the climatic and socio-economic changes within cities. Furthermore, the uncertainty in the future climatic and socio-economic conditions poses significant challenges for planners. A framework is proposed for analysing urban systems with evidence-based tools over extended time scales. This forms the basis of a manifesto for future challenges and research directions for this critical subject area, which ultimately will help engineers and urban planners to better understand the areas for which they are responsible and to develop adaptation strategies that can tackle the challenges posed by long-term global change and lead to more sustainable cities.
1. Cities on the front line
Urbanization is one of the most powerful and visible anthropogenic forces on Earth. The twentieth century has seen humans migrate from being rural to urban species. Cities occupy less than 3% of the Earth's land surface (Balk et al. 2005) but house over 50% of the world's population, a figure that was only 14% in 1900 (Douglas 1994) and is estimated to be 60% by 2030 (UN 2004a). The rate of growth in developing countries is still faster (UN 2004b).
Cities are remarkable in that they are manifestations of coupled cultural and technological systems but also potential hot spots of climate vulnerability due to the high concentration of people and infrastructure. These are real and urgent problems: the European heat wave of 2003 resulted in 35 000 deaths primarily in urban areas (Fink et al. 2004), and floods in Europe in 2002, 2005 and 2006 have led to the loss of hundreds of lives and cost billions of pounds, while annual damages associated with these and other climatological events are rising (Mileti 1999).
However, the influence of urbanization extends far beyond the palpable terraforming that occurs within urban boundaries. Resources consumed by city dwellers result in land-use changes and resource movements between other rural and urban areas that extend far beyond the physical or political urban boundaries. Urban activities release greenhouse gases (GHGs) that drive global climate change directly (e.g. petrol-based transport) and indirectly (e.g. electricity use and consumption of industrial and agricultural products). As much as 80% of global GHG emissions are estimated to be attributable to urban areas (MunichRe 2004).
Conversely, cities therefore represent concentrated opportunities for adaptation to climate impacts and mitigation of GHG emissions. However, this involves complex interactions of citizens, governmental/non-governmental organizations and businesses. This complexity can inhibit the development of integrated strategies (which may involve transportation demand management, land-use planning and construction of new civil infrastructure) whose combined effect is more beneficial than the achievements of any single agency or organization acting unilaterally.
While the role of policy-makers and scientists in addressing these challenges is not underestimated, it is engineers who will be at the forefront of delivering adaptation technologies and strategies. Nowhere is this more relevant than in cities where they have a pivotal role to play in contributing to the urgent challenge of reducing GHG emissions and adapting our built environment and infrastructure systems to be resilient to the impacts of climate change.
Following this introduction, this paper provides an overview of the challenges posed to cities by climate change, before considering the difficulties of designing adaptation strategies for cities under uncertain future conditions. A framework is proposed for addressing urban-scale analysis of climate impacts which forms the basis of a manifesto for future challenges and research directions.
2. Climate change and cities
Globally, climate change is expected to lead to an increase in mean sea level and changes in frequency, intensity and spatial patterns of temperature, precipitation and other meteorological phenomena, such as wind and cloud cover (IPCC 2007a). The type of impacts, their directionality and magnitude will differ for each city, but some key impacts are identified (DoH 2001; Nicholls 2002; Sanders & Phillipson 2003; Dawson et al. 2006; Wilby & Perry 2006; IPCC 2007b).
Sea-level rise will increase the risk of storm-surge flooding and rates of coastal erosion thereby threatening beaches, coastal settlements and wetlands. Raised sea levels can also lead to saline intrusion into freshwater aquifers and impeded drainage of extreme flows in urban drainage systems and rivers.
Flood risk will be altered by changes in precipitation patterns. For the majority of cities, the number of intense precipitation events is expected to increase and place additional strain on fluvial and urban flood systems, leading to increased disruption for businesses and inhabitants.
Buildings and infrastructure are likely to experience increased damage from more frequent and intense windstorms and other extreme weather. Groundwater movement and changes to temperature and precipitation patterns are likely to lead to increased incidence of damage from subsidence, heave and landslides. Transport networks are likely to experience increased disruption from more frequent extreme weather events, while temperature rises will reduce passenger comfort and increase damage to infrastructure (e.g. buckled rails) but reduce cold-related disruption. Solar and wind energy systems may benefit from increased solar radiation and windiness, but increased temperature runs the risk of the energy distribution network overheating. Increased temperature, coupled with water shortages (for cooling), and increased demand for air conditioning could strain energy generation in heatwaves.
Droughts are expected to increase in frequency over many areas, with implications for water resources in terms of both quality (and concomitant implications for health and aquatic ecosystems) and availability for human consumption, industry and neighbouring agricultural areas.
Health effects are likely to see reduced cold-related deaths, but more heat-stress deaths. Higher UV exposure (from reduced cloud cover) will lead to increased incidences of cataracts and skin cancer, while higher temperatures and more still days encourage photochemical smog and ozone. Warming and reduced water quality have potential to facilitate migration and establishment of vector and water-borne diseases in new regions and increase incidences of food poisoning. More frequent extreme weather is likely to increase incidences of weather-related injury and loss of life.
Business and the urban economy are likely to be affected by lifestyle changes of city inhabitants and tourists (e.g. warmer climate may lead to more outside activity). Consumer demand and behaviour may change, affecting industries that may also need to alter their operating procedures (e.g. warmer winters provide better conditions for construction). The insurance industry will experience increased exposure to weather-related risks.
Biodiversity and urban ecology will be influenced by changes in the temperature and precipitation regime, resulting in exotic species (including pathogens and pests) establishing themselves in new areas. Furthermore, the phenology (e.g. flowering period, migration patterns) of plants and animals will change. Sea-level rise will squeeze intertidal habitats and salt marshes.
Resources outside the urban boundary will be increasingly stressed by water shortages, wildfire, wind damage to crops and increased ocean temperature changing fish stock abundance. Waste management has implications in terms of GHG emissions generated (e.g. from landfills, processing and transportation), energy generation and more generally in the context of sustainability and resource use. Many of these are likely to be affected by climate change.
Many impacts can be further aggravated through systemic interactions with the built environment and socio-economic pressures (e.g. warmer weather placing further demand on personal and agricultural water resources). Likewise, the interconnectivity of infrastructure systems can lead to cascading failures when subjected to extreme weather events, for example inundation of water treatment works or power stations can cause disruption disproportionate to the extent of the flood.
(a) Systemic climate–city interactions
Urban areas interact with local climatic conditions, frequently amplifying their impacts. It has been recognized for many years that urban areas generate heat islands (Howard 1818). These are caused by the storage of solar energy in the urban fabric during the day and release of this energy into the atmosphere at night: the process of urbanization replaces the cooling effect of vegetated surfaces by imperviously engineered surfaces with different thermal properties (Oke 1982). Furthermore, anthropogenic sources (e.g. central heating systems, air conditioning, transport, industrial processes) emit heat directly into the urban area, while buildings and infrastructure increase surface roughness that can reduce wind speeds, convective heat loss and evapotranspiration.
Warm, still days reduce air quality because high temperatures and ultraviolet light stimulate the production of photochemical smog, ozone and other compounds from traffic and industrial emissions and plants (Kovats et al. 1999; DoH 2001).
High-density cities use significantly less energy per capita on private transport (Newman & Kenworthy 1999), but generate more intense urban heat islands and aggravate other issues, such as flooding and subsidence. Furthermore, densification of cities can lead to a loss in quality of life for many residents (Heath 2001).
At different scales, different components of the urban system become important: building materials have different thermal properties and subsequent implications for the heat island and roofs can influence airflow locally while the configuration of buildings and infrastructure within the wider urban area has implications for other impacts, such as wind and heat fluxes (Blankenstein & Kuttler 2004), flood risk (Dawson & Hall 2006) and (waste) water management (Buxton 2006).
A high concentration of population and buildings does not necessarily imply significant climate impacts as vulnerability is also a function of social, economic and political processes (Adger & Vincent 2005; Allenby & Fink 2005). This includes factors such as the cities' economy, population demography, institutional stability and corruption, global interconnectivity, dependence on natural resources and public infrastructure. Measures to reduce vulnerability might include the diversification of ecological and economic systems and building inclusive governance structures—essentially taking a portfolio approach to minimizing risks across society in the broadest sense.
(c) The urban footprint
The majority of settlements are not isolated and interact strongly with neighbouring areas, and increasingly the rest of the world. This interaction occurs through a complex network of flows of energy, transport, materials, food, waste and water. Factors such as increasing global population or increased demand for resources to accommodate lifestyle changes have changed, and will continue to change, these flows (Pugh 1996). For example, demand for meat in Mexico City in the mid-twentieth century led to deforestation to provide cattle grazing areas 400 km away (Barkin & Zavala 1978). In addition, raw materials, design, production and selling of manufactured goods increasingly occur in multiple locations, leading to increased transportation emissions and displacement of industrial emissions from the end-user of the product. And efforts to reduce GHG emissions have seen food crops replaced by more profitable biofuels in agricultural areas (Cassman 2007). Many of these flows, particularly those related to agriculture, are susceptible to climate impacts, such as land degradation, salinization of aquifers, soil erosion and changed crop yields (although yields could be increased in many areas). Industries (including energy generation) are also vulnerable to water shortages and other climate hazards.
3. Adaptation to global change
(a) Adaptation versus mitigation
Responses to climate change in cities are aimed at reducing net GHG emissions (mitigation), and at the impact of climate change through adjustments to social, natural or built systems (adaptation). Benefits of mitigation in terms of climate impacts reduction tend to be experienced globally over longer time scales (although local improvements in air quality are observed in shorter time scales). Adaptation tends to provide regional and local reductions to climate impacts while also reducing vulnerability to natural variability in weather (figure 1).
Adaptation and mitigation can be implemented at a wide range of scales. Households may mitigate through installation of energy-efficient devices or adapt through floodproofing, while nationally carbon taxes and building codes seek to achieve similar effects on a larger scale. Other actors such as the insurance and finance industries can influence planning policy by exacting higher insurance premiums or withdrawing financial support altogether where risks are considered too high. Arguably, it is at the scale of cities and regions where mitigation efforts have been most effective in terms of implementing practical strategies and lobbying governments for legislation (Kousky & Schneider 2003).
Adaptation is not a new activity, humans have always been adapting to their environment, often unintentionally. If poorly managed, or not considered at a broad scale and with a view to long-term consequences, climate change can induce energy-intensive maladaptations, such as air conditioning, pumped drainage or desalination. These energy-intensive adaptations can undermine efforts aimed at mitigating GHG emissions. Moreover, failure to consider a range of possible impacts over extended time scales can lead to undesirable ‘lock-in’ to specific adaptation options (Brewer & Stern 2005). For example, construction of flood defence infrastructure can lead to intensive flood plain development that subsequently ties flood plain managers to further flood defence infrastructure (Kates 1971) as alternatives such as managed retreat become prohibitively expensive.
Many of the adaptation technologies, and methods used to appraise their effectiveness, have evolved from the management of extreme events. However, there are important distinctions between adapting to climate change and managing extreme events. The climate is now very different from the conditions under which most cities were founded and is changing at a previously unanticipated rate. Development and planning has previously assumed a stationary climate, a position which is no longer tenable. Furthermore, climate change transcends multiple disciplines and persists over long time scales—existing tools are not always well suited to this type of analysis. However, our improved scientific understanding, modelling and monitoring capabilities now provide the opportunity for developing evidence-based tools to support the planning and implementation of adaptation strategies.
(b) Adaptation under uncertain future conditions
Uncertainty analysis involves the systematic identification and quantification of the sources of uncertainty and their potential implications in order to understand how predictions of interest to a decision-maker may vary under plausible variations in the assumptions made in an analysis. The size and influence of cities are in continual flux, driven by climate change and many other factors (Turchin 2003). This adds further difficulties to appraising adaptation options as they are inevitably subjected to both climatic and socio-economic uncertainties.
Probabilistic analysis has long been established in engineering decision-making. The climate science community is moving in the direction of probabilistic scenarios to quantify future uncertainties (usually conditional on a particular GHG emissions trajectory). Although Hall et al. (2007) warn against reducing climate variables to single distributions in instances that clearly misrepresent scientific disagreement, they also propose the use of imprecise probability distributions that avoid decision-makers identifying apparently optimal adaptation decisions through naive idealization of the uncertainties. Sensitivity analysis (Saltelli et al. 2000) extends traditional probabilistic analysis by apportioning the contribution that model input variables, acting independently or in combination, make to the uncertainty in output quantities of interest. Sensitivity analysis therefore provides a rational justification for investment in data collection or further studies.
Methods for dealing with socio-economic scenarios vary more widely and tend to be more discursive. Effective urban management requires combined use of climate change and socio-economic scenarios. Inevitably, it is not possible to predict all future effects of an adaptation strategy (Collingridge 1980), but a robust system will perform reasonably well even in situations which depart considerably from expectations. One quantitative theory of robustness is info-gap analysis (Ben-Haim 2001), which can be used to identify how the performance of different adaptation strategies deviates as conditions depart increasingly from expectations. However, there may be a trade-off between efficiency under assumed conditions and robustness.
Once decision-makers have taken reasonable measures to understand the implications of key uncertainties, and reduce them where it is feasible to do so, they should be able to proceed with making choices that are as far as possible robust to uncertainties. Exploring the variable space (and its subsequent impact on decisions) as broadly as possible provides an alternative to—or can be used to complement—conventional cost–benefit analyses (Lempert et al. 2003) for ranking alternative adaptation strategies and a basis for constructing robust portfolios of climate mitigation and adaptation measures.
(c) Design of adaptation strategies
Evans et al. (2004) identified 86 adaptation options in flood risk management alone; consideration of all possible adaptation options is impractical in this brief paper. However, it is possible to identify some generic modes of adaptation, which include engineering infrastructure (e.g. reservoir construction), increasing the resilience of natural systems (e.g. restoration of salt marshes), reducing impacts in the built environment (e.g. land-use planning, building codes), reducing vulnerability (e.g. education programmes, governance), risk transference (e.g. insurance), monitoring (e.g. remote sensing) and emergency management (e.g. warning systems, evacuation planning). Desirable features of an adaptation strategy therefore include the following.
Taking a broad definition of the urban system. Adaptation may have an impact on any part of the urban environment, and adaptation strategies should be considered in the context of the physical processes, human-built infrastructure, economic, social and environmental systems that are impacted on, and/or influence climate change processes. Opportunities for adaptation should not be constrained to the geographical or administrative boundary of the city. For example, upstream catchment storage can reduce the probability of flooding within the city boundaries.
Continuous management. The urban system is dynamic. As such, urban policy-makers need to engage in a constant ‘dialogue’ with the system, collecting information about its behaviour and, in parallel, responding to this monitoring by taking appropriate intervention actions. Designing for specific ‘events’ should be replaced by consideration of the performance of the urban system over a full range of behaviours, future outcomes and management strategies. To enhance the robustness of an adaptation strategy, it is necessary to ensure that any negative effects are detected as early as possible and the adaptation strategy remains controllable. This requires regular monitoring, reducing the costs of negative impacts and the response time to correcting them, avoiding lock-in through use of flexible responses and employing a diverse range of options.
Tiered analysis and iterative planning and decision-making. Adaptation to climate change within cities cascades from high-level policy decisions (e.g. planning policy), based on coarse broad-scale analyses, to individual designs and projects, which require more detailed analysis. High-level policy and plans provide the framework and common understanding within which more detailed actions are implemented. Adaptation should be timed to coincide with natural planning points where it is possible (e.g. scheduled refurbishing or new developments) to minimize costs.
Developing portfolios of options. A portfolio approach seeking to manage multiple climate impacts across the urban system, by combining adaptation options from the range of modes identified previously, will be most effective. Wherever feasible win–win (i.e. benefits across multiple sectors) or no-regret approaches should be sought, although there will frequently be tensions and competing objectives. Development of integrated portfolios requires coordinating the activities of more than one organization and multiple stakeholders.
The political and administrative contexts. Managing climate change impacts at a city scale will remain an abstract concept unless placed within the current policy and administrative contexts. This involves using existing policy instruments and actively identifying opportunities to influence policy change. It may involve reacting opportunistically to policy, administrative or regulatory reviews and changes that are initiated for non-climate-related reasons.
4. Towards integrated urban analysis
Clearly, on the scale of cities it is meaningful to think about climate impacts, adaptation, mitigation and sustainability in the same quantified assessment framework. Moreover, cities can hardly be divorced from their regional and global contexts. Analysing urban systems with evidence-based assessment tools can help cities escape from the vicious circle of increasing climate impacts and emissions. Evidently, an integrated assessment of urban areas (figure 2) should involve:
quantitative evaluation of a wide range of climate impacts, GHG emissions and other resource flows;
framing city scenarios and impacts analyses within the context of global climate and socio-economic change;
analysis over the extended temporal and spatial scales that are relevant to urban policy-makers addressing the challenges posed by climate change;
capturing the interactions and feedbacks between economy, land use, climate impacts, GHG emissions, resource flows and broader issues of sustainability;
analysis of both adaptation and mitigation options that can be implemented at a range of scales (e.g. from buildings through to national planning policy);
facilitating the construction of multi-sector portfolios of management options and testing their robustness under a wide range of possible future outcomes; and
use of appropriate visualization and stakeholder participation methods to ensure effective communication of information between policy-makers, scientists and members of the public.
Realization of the facility envisioned in figure 2 is not a small task. There has been substantial research in many of the areas considered previously, but significantly less from an integrated perspective. Some key challenges are now considered.
Climate impacts analysis in urban areas will require the development of weather generators of increased resolution in space and time, capturing both variability and spatial correlation between different types of weather for analysis of multiple climate impacts. This is naturally extended to analyse cascading failure of infrastructure (and other) components in the urban system. Notably, this analysis should consider implications of changes to the energy supply system such as increased deployment of decentralized and renewable energies, in the context of its robustness and reliability under changing climatic conditions. Remotely sensed data should be used to facilitate automation of model parametrization (e.g. identification of roads and embankments). Furthermore, some impacts may require the development of new modelling paradigms. For example, modelling of urban drainage and air quality at a broad scale involves computationally expensive simulation of interactions between flow and local features. If these types of models are to be included in risk and uncertainty analyses at an urban scale other than for testing a limited set of simulations, development of emulators (see Mayer et al. 2000), statistical methods or alternative modelling approaches may be necessary. An important consideration, in these cases, is whether urban-scale modelling provides benefits in proportion to the limitations imposed by the additional computational expense.
Improved understanding of urban function and dynamics, particularly the multifunctional aspects and relationship with external regions and drivers, will require better understanding of how global and national drivers (e.g. economic, social, technological, climatic change) influence urban change and vice versa. Wider interactions and feedbacks that deserve further consideration include: (i) interactions within the urban area between land use, travel patterns, public and private transport infrastructure and employment and population demography, (ii) the impacts of natural and man-made hazards, climate changes and feedbacks from adaptation and mitigation strategies, and (iii) the interaction between urban areas and changes outside their borders.
Analysis of sustainability requires identification and monitoring of appropriate metrics. These should incorporate, among others, measures such as social justice, environmental issues, liveability and robustness. This should build on existing work on analysing the urban footprint (see BFF 2002) to improve accounting of embedded energy in materials and resources that have been manufactured and processed outside the urban boundary and to better understand the relationship between waste handling, recycling, processing, landfill processes and emissions. It is important not to focus on a single metric (e.g. CO2 emissions) as this may lead to unforeseen increases in other environmental impacts.
Integrating technologies are required to assist in the coupling of databases and models. Currently, this requires the ‘patching’ together of models in a bespoke manner which is a barrier to a longer term aim of more routine integrated urban assessment. Not least among the challenges are the software issues of integration and the commercial issues associated with modularity and standardization. Furthermore, it is necessary to design integrated modelling frameworks that are useful for constructing sets of simulations and outputs for uncertainty analysis. To address the range of scales at which urban systems are managed and that physical processes manifest themselves demands a tiered and nested approach to modelling which integrates a range of low- through high-complexity models operating at a range of spatial and temporal scales.
Decision-support and visualization. Effective decision-making relies upon the engagement of stakeholders, of which there are many in urban areas. New tools for visualization are providing a common platform for communication and negotiation. For example, a virtual reality ‘decision-theatre’ is being pioneered in Phoenix (USA) to support stakeholder engagement and evidence-based decision-making (DCDC 2006). Research is needed in order to develop appropriate tools to support long-term decision-making for a range of stakeholders that also communicate climate impacts, uncertainties, sustainability measures and the wider interactions of a city. These tools and methods must be able to assimilate large quantities of evidence from a wide range of sources.
A long-term urban research programme. A multitude of organizations collect vast amounts of data at varying frequencies and resolutions for a diverse set of economic, social, physical and environmental attributes of urban systems. Data quality is improving as remote sensing and other monitoring techniques are becoming more accurate and densely deployed. Ecological research in the USA has benefited from structured, place-based research programmes (Grimm et al. 2000). A similar programme focusing on climate and socio-economic change in the urban context could provide an unparalleled data repository and resource for urban research.
Cities are major drivers of both global climate change and hot spots of potential impacts to climate-related hazards. It is no longer tenable to consider urban systems to be static artefacts constructed in a stable environment. The environment within which urban systems are constructed is changing for a variety of reasons, including long-term climate and socio-economic changes. Management of cities therefore becomes a process of dynamic control under conditions of uncertainty, and to paraphrase Charles Darwin: it is not the strongest of the cities that will survive, but rather the ones most responsive to change.
Cities can most effectively respond to climate change while meeting their sustainability objectives, through integrated analysis of social, economic and environmental factors to support the assembly of portfolios of adaptation measures—a framework for which has been proposed here. These portfolios will include conventional engineering approaches as well as other measures such as planning and financial instruments. Despite significant advances in both integrated assessment tools and climate impacts analysis, challenges remain before this can become routine. However, development of this new capacity for integrated assessment of cities over an extended time scale will help city planners and decision-makers develop and implement better plans and strategies to adapt to global changes in urban areas. This will lead to more sustainable and efficient cities.
R.D. is funded by the Tyndall Centre for Climate Change Research Cities Programme.
One contribution of 20 to a Triennial Issue ‘Chemistry and engineering’.
- © 2007 The Royal Society