IvIEWPOINT Climate change and climate impacts Please don’t confuse the two!
Stewart J. Cohen
Stewart J. Cohen is with the Canadian Climate Centre, Downsview, 5T4, Canada.
The author would like to thank Ian Burton, Barrie Maxwell, Yongyuan Yin and Jamie Smith for their comments. Opinions expressed are those of the author and do not necessarily reflect the views of his COIleagues or Environment Canada.
It is worthwhile to study future cliIt is worthwhile to study the impacts of future climate? Are there other comparable fields of research where people work on consequences or impacts of ‘futures’ without knowing much about rates or scales of change? Governments and private sector often work with ‘pictures’ of the future, or ‘scenarios’. Investment decisions, construction of infrastructure, land-use plans, trading agreements and resource management goals all require consideration of the long-term view, despite the uncertainties in projecting any kind of trend. Although previous experience is important, the future cannot be described by simple extrapolations from the past. One of the futures that is of concern is a projected increase in global mean temperature (‘global warming’), which might be anticipated as a result of increased atmospheric concentrations of CO1 and other trace gases (ie enhanced Greenhouse Effect). This has received considerable attention in recent years from the scientific and policy communities. General Circulation Model (GCM) simulations have played an important role in shaping the scientific debate as well as the mate?
1993 Federal Government
development of international policy instruments, particularly the Framework Convention on Climate Change which was signed by approximately 150 countries at the 1992 United Nations Conference on Environment and Development (UNCED). The UNCED document focuses largely on the need to limit the growth in atmospheric concentrations of trace gases. Much less attention has been given to response options that may be needed if limitation is not successful. Such options are generally called ‘adaptation’ strategies. Within this set, there are some known as ‘no regrets’ options that could be implemented without climate change. Examples include the promotion of energyefficient technologies and droughtresistant crops. Without knowledge of potential impacts of climate change, however, other possible adaptation strategies will be difficult to identify because governments and the private sector will not know what they might be adapting to.
The two challenges: why are they separate? Information about projected global warming has come largely from GCM
of the USA (text) & Butterworth-Heinemann
Ltd (layout & typography)
‘J.T. Houghton, G.J. Jenkins and J.J. Ephraums, eds, Climate Change: The If CC Scientific Assessment, World Meteorological Organization/United Nations Environment Programme, Cambridge University Press, Cambridge, 1990; J.T. Houohton, B.A. Callander and SK. Varney, eds, Climate Change 7992: The Supplementary Report to the /PCC Scientific Assessment World Meteorological Organization/United Nations Environment Programme, Cambridge University Press, Cambridge, 1992. ‘R.E. Munn, ‘The framework for a climate impact assessment’, in Carbon Dioxide Issues and Impacts: Proceedings of the Workshop on Carbon Dioxide issues and Impacts, Toronto, 28-29 August 1979, Atmospheric Environment Service, Downsview, Canada, 1979; R.W. Kates, J.H. Ausubel and M. Berberian, eds, Climate impact Assessmen& SCOPE 27, John Wiley, New York, 1985. 3M.H. Glantz, R. Katz and M. Krenz, eds, Climate Crisis: The Societal impacts Associated with the 1982-83 World- Wide Climate Anomalies, United Nations Environment Programme, Nairobi, and the National Center for Atmospheric Research, Boulder, CO, 1987. 4W.E. Riebsame, Assessing the Social Implications of Climate Fluctuations: A Guide to Climate impact Sfudies, United Nations Environment Programme, Nairobi, 1988. 5For example, see inadvertent Climate Modification: Report of the Study of Man’s impact on Climate (SMIC), MIT Press, Cambridge, MA, 1971. 6W.L. Gates, ‘Numerical modelling of climate change: A review of problems and prospects’, in Proceedings of the WMO/ IAMAP Symposium on Long-Term C/mate Fluctuation, WMO No 421, World Meteorological Organization, Geneva, 1975, pp 343-354. ‘S. Arrhenius, ‘The influence of the carbonic acid in the air upon the temperature of the ground’, Philosophical Magazine, Vol 41, 1896, pp 237-276. ‘S. Manabe and R.W. Wetherald, ‘The effects of doubling the CO, concentration on the climate of a general circulation model’, Journal of Atmospheric Sciences, Vol32, 1975, pp 3-15. ‘S. Seidel and D. Keyes, Can We Delay a Greenhouse Warming?, US Environmental Protection Agency, Washington, DC, 1983. Other important documents from that period include: W. Bach, J. Pankrath and J. Williams, eds, lnteractions of Energy and Climate, Reidel, Dordrecht, 1980; and Proceedings of the World Climate Conference, WMO No 537, World Meteorological Organization, Geneva, 1979.
simulations. There is much uncertainty associated with GCMs of the atmosphere, particularly with regard to their ability to project regional-scale changes. There appears to be consensus regarding the principle of COZinduced global warming, though debate continues about its rate and magnitude, which could be unprecedented.’ The possibility of climate change has led to concern about potential impacts on natural and social systems at global and regional scales. A relatively new field of study has emerged, commonly known as ‘climate impact assessment’.’ Besides the global warming issue, climate impact assessment has also been applied to historical events, including El Nitio3 and a wide range of climatic extremes such as drought.’ Policy and research questions associated with global warming are both multidisciplinary and interdisciplinary, and involve both physical and social sciences. However, it is important to realize that two separate challenges are being addressed - climate change and climate impacts. The former centres on projections of the climate itself (using GCMs, hypothetical cases, historical analogues, etc), though the projection of the timing of climate change is also dependent on projections of energy consumption and emissions. The latter uses climate projections as scenarios (not forecasts) to serve as inputs to another set of methodologies for which there are other uncertainties and constraints. For example, how can the impacts of climate change on agriculture be described if there are uncertainties in projections of climate, prices, technology and trade patterns? It appears that the two activities have often been treated as one by scientists, politicians, policy advisers, and the public. Issues related to impacts research may often be masked and distorted by the debate about global warming. Often, the impacts researcher is asked questions about climate models rather than climate impacts methodologies. It is easy to forget that impacts researchers are probably not climate modellers. The misperception creates confusion within the user community (including
stakeholders and the public in general) and may discourage participation by users beyond that of being spectators and critics. In my view, climate change and climate impacts are indeed separate issues and must be seen as such. Otherwise, research on climate impacts will be viewed as only a subset of climate change research, rather than an activity with its own methodological challenges, which could be applied to various climate-related problems. Research on climate change and climate impacts poses different challenges and addresses different research questions. Spectators, whether they are scientists or stakeholders, should understand this fundamental difference. It is also important to be aware of the methodological concerns that are being faced by impacts researchers, and the need to maintain a serious long-term effort in this complex interdisciplinary field of study, regardless of the status of GCMs or the debate about trace gas limitations.
The two challenges: research with different origins Although most people now associate climate modelling almost exclusively with the subject of global warming, modelling began long before the current debate about warming. GCMs were first developed in the 1960~~ Model validation and intercomparisons were documented in the 1970s more than a decade before the creation of the Intergovernmental Panel on Climate Change (IPCC). The possibility of global warming as an outcome of increased atmospheric concentration of CO2 and other greenhouse gases had been identified by the Swedish scientist Arrhenius as early as 1896,’ and by climate modellers in the 1970~s Global warming has been a political issue, however, only since the 198Os, particularly after the publication of the US Environmental Protection Agency’s 1983 report Can we delay a greenhouse warming?.’ What had been a scientific question suddenly became the focus of growing political activity, focusing for the most part on strategies to limit the increase of
“I. Burton, R.W. Kates and G.F. White, The Environment as Hazard, Oxford University Press, New York, 1978; H.H. Lamb, Climate, History and the Modern World, Methuen, London, 1982. “Kates eta/, op tit, Ref 2; World Meteorological Organization, op cif, Ref 9. “World Meteorological Organization, The Changing Atmosp‘here: /&plications for Global Securifv. WMO No 710. World Meteorological ‘Organization, Geneva, 1988. 13For example, coarse resolution, the role of clouds, and difficulties in modelling ground surface processes.
CO1 and other radiatively active trace gases. Adaptation strategies did not receive the same attention. Perhaps this was because of insufficient information about the costs of adaptation compared with the costs of doing nothing. It may also have been due to the possibility that some economic sectors or geographic regions could benefit from a warmer climate. Climate impact assessment owes much of its origins to studies of natural hazards, and climate and history, both of which focused on observed extreme events.“’ The first World Climate Conference, held in 1979, was a response to a series of extreme events, including the reduction of precipitation in the African Sahel, widespread desertification in sub-Sahara11 Africa, and a perception that climate variability (rather than just temperature) was increasing. It was also argued that societal vulnerability was increasing, and that CO*-induced climatic changes could lead to sea-level rise, expansion of deserts, and other disruptive effects in the 21st century.” Scientists and governments organized a series of meetings about global warming during the latter half of the 1980s. One of these was the World Conference on the Changing Atmosphere: Implications for Global Security, held in Toronto in 1988.” This stimulated the formation of the IPCC and the drafting of the Framework Convention on Climate Change within four years - an unprecedented rate of progress in international environmental negotiations. Climate change research and climate impact assessment have always been subject to serious scientific criticism, particularly regarding uncertainties in methodology and input data. In recent years, however, there has also been mounting criticism of a political kind. The rapid pace of the negotiations for the Framework Convention alarmed those who were concerned about the policy implications of CO1 limitation strategies, particularly carbon taxes and controls on deforestation, These political critics found allies within the scientific community, and so we are now entering a period of ‘global warming bashing’. This includes ‘GCM bashing’ in particular,
but also attacks on impacts studies that use GCM outputs as the bases for ‘predictions’. A number of scientists have been drawn into the political aspects of the debate, and have found themselves subjected to political attacks.
Impacts research in the 1990s The political debate between those who are for limitation of greenhouse gas build-up and those who are against it, has spilled over into the arena of climate impacts research. As a result, the important tasks of improving impacts methodologies and communicating impacts information to potential users have been sidetracked. Now that the scientific question of global warming has acquired significant political dimensions, it is important to underscore the fact that climate modellers and climate impacts researchers have different backgrounds and different scientific missions. Climate modellers are generally atmospheric scientists and oceanographers. Impacts researchers include physical, biological, and social scientists who often work at smaller scales of time and space than the specialists who build GCMs. Climate modellers seek to improve knowledge of the climate system. Impacts researchers focus on linkages between climate and various other systems that may be climatesensitive (eg water resources, agriculture, regional economies). The shortcomings of GCMs have become well known13 but impacts modelling has its own - quite different - methodological problems. Impacts research poses a specific challenge that deserves attention in its own right. What is the nature of this challenge? It is to build a bridge between climate and regional resource issues by accounting for many factors that could shape regional futures, while keeping both issues in their proper contexts. Why should we care about impacts? Without knowledge about impacts, stakeholders will not know the costs of doing nothing. The first impacts case studies of global warming scenarios that used GCM-based data were initiated during the mid-1980s. These were single-
‘“N.J. Rosenberg et al, Processes for Identifying Regional Influences of and Responses to Increasing Atmospheric CO, and Climate Change - The MINK Project, Office of Energy Research, Carbon Dioxide Research Program, TRQ52, United States Department of Energy, Washington, DC, 1991. 15M.L. Parry, T.R. Carter and N.T. Konijn, eds, Assessment of Climate tmpacts on Agriculture: Vol 7, Assessments in Cool Temperate and Co/d Regions, Reidel, Dordrecht, 1987. 16W.J. McG. Tegart, G.W. Sheldon and D.C. Griffiths, eds, Climate Change: The /PCC impacts Assessment, World Meteorological Organization/United Nations Environment Programme, Australian Government Publishing Service, Canberra, 1990. “Op cit. Ref 14. ‘%:J. Cohen, ‘Impacts of global warming in an Arctic watershed’, Canadian Water Resources Journal, Vol 17, No 1, 1992, pp 55-62; T.R. Carter, M.L. Parry, S. Nishioka and H. Harasawa, Preliminary Guidelines for Assessing impacts of Climate Change, Environmental Change Unit, Oxford, and Center for Global Environmental Research, Tsukuba, 1992. “IPCC, 1990, op tit, Ref 1. 2oIPCC, 1992, op tit, Ref 1; T.R. Karl, R.R. Heim, Jr and R.G. Quayle, ‘The greenhouse effect in Central North America: If not now, when?‘, Science, Vol 251, 1991, pp 1056-l 061.
sector studies, often based on only one GCM. In recent years, impacts research has expanded to include multiple sectors for regions” and global coverage of individual sectors, most notably agriculture.‘” National studies have been completed in New Zealand, South-east Asia, the USA, the UK and Australia, and the IPCC has published a worldwide literature review which will soon be updated.” A number of studies have used more than one scenario of climate change, and some are including scenarios of population and technological changes so that it is possible to move beyond the ceteris paribus assumption. During the past ten years, impacts researchers have encountered a variety of difficulties including: .projecting impacts of extreme events, applying models to scenarios that may be beyond the range of model calibration, and providing estimates of probabilities and errors. Many of these difficulties stem from uncertainties about climate change itself. Moreover, the quantification of impacts, particularly in economic terms, is hampered by poor understanding of modelling of climate-society linkages. This is an important challenge to impacts research, regardless of the progress that is made in understanding climate change. Besides uncertainties in GCMs, it is likely that impacts studies in the 1990s will continue to be hampered by inconsistent methodology for scenario development, data gaps. and incomplete knowledge of linkages between climate and other systems, including human systems. Besides these methodological challenges, there are several important reasons for maintaining - and indeed expanding - regional-scale impacts research programmes, regardless of the status of GCMs or greenhouse gas limitation policies. These include: 0
Linkage between global warming science and regional policy concerns could be quantified. 0 Scientific uncertainties could be expressed in terms of impacts, not just temperatures. 0 The regional focus would encourage development of interdisciplinary and interjurisdictional col-
laborations and partnerships, because the various parties would be attracted by their common interest in the region under consideration, and each would gain some ownership in the results of the exercise.
What then can be expected from impacts research in the 199Os, given the politicized nature of global warming and the continuing scientific uncertainties? The recently completed MINK study” has come closest to producing an integrated regional assessment of a global warming scenario, and there will be more such attempts to describe the indirect but potentially significant linkages between climate and regions, and to provide some guidelines on impact assessment methods.‘s Even singlesector studies (eg water resources, agriculture, urban centres) will become more sophisticated and complex as they strive to account for scientific, management and policy-related concerns. Impacts research may, however, still provide a stage for some who are more interested in current rather than future issues. Some have argued that impacts studies of climate change should wait for improvements in GCMs, reduction of uncertainties, or a clear signal that global warming has begun. There are indications, however, that more than a decade will elapse before scientific uncertainties will diminish’” or before a global warming ‘signal’ can be clearly detected.2” Meanwhile, long-term resource management and planning options are being considered by governments and industries now, with information that is available today, but information that is often incomplete. These options include water resources agreements, afforestation and transportation plans. Some of these involve considerable investments of time and money. A number of no-regrets’ options could be considered (eg development of a drought management plan), but there will continue to be uncertainty in identifying the optimum adaptation options if there is insufficient information about what regions should be adapting to. For this, it will
be necessary to have better climate models, more environmental and socioeconomic data, and better assessments of impacts of projected changes in climate. Improved GCMs by themselves will not reduce uncertainties related to adaptation. Despite the uncertainties, it would be prudent to generate, review and
publish information about potential impacts of climate change scenarios so that interested parties could have access to it, caveats and all. For anyone who believes in reducing uncertainties about global warming and its implications for our planet, a continued interdisciplinary effort is really the only alternative available.