Preventing Dangerous Climate Change

Preventing Dangerous Climate Change Adaptive Decision-making and Cooperative Management in Long-term Climate Policy Jürgen Scheffran, Ph.D. Adjunct Associate Professor Political Science and Atmospheric Sciences ACDIS, University of Illinois 359 Armory Building, MC 533, 505 East Armory Ave. Champaign, IL 61820, USA Phone 217-244-0463 Email [email protected]

Semi-final draft of a paper published in: Velma I. Grover (ed.), Global Warming and Climate Change - Ten Years after Kyoto and Still Counting, Science Publisher, 2008, pp. 449-482.

Abstract: With the entry into force of the Kyoto Protocol to the UN Framework Convention on Climate Change (UNFCCC) the international community moved one step forward in addressing the problem of global warming. The Kyoto instruments, such as the clean development mechanism, joint implementation and emissions trading, are important to strengthen international cooperation. A lack of agreement on the causes, implications and dangers of climate change impedes progress on appropriate actions for mitigation and adaptation to achieve the ultimate objective of Art. 2 UNFCCC of preventing “dangerous anthropogenic interference with the climate system”, including its specifications with regard to timing, ecosystem adaptation, food security and sustainable economic development. Comprehensive approaches of integrated assessment take into consideration decision-making and negotiations among multiple levels and actors, based on risk perceptions and evaluations, complexities and uncertainties, vulnerabilities and adaptive capacities as well as critical thresholds of abrupt climate change. Value judgements of stakeholders are translated into admissible domains of the climate system and actions for emission reductions, energy investments and carbon capturing and sequestration. Identifying indicators for dangerous climate change and effective implementation of policies are important to avoid or minimize risks. Joint action of climate coalitions is essential to achieve the benefits of cooperation.

1

Introduction Imagine a boat with a number of passengers, each having a paddle, floating on a wild river through a rugged and foggy landscape with potential but unknown dangers. Passengers have only a limited perception of the landscape and use the paddle whenever the boat seems to move too fast towards a potential danger area. Some passengers may be too weak or lazy to take any action or they do not care about the dangers, while others try to get through the danger area as fast as possible to reach their destination. Some are cautiously trying to slow down the speed of the boat to keep it under control, and a fourth group wants to reach a safe harbor to make a plan or wait for better circumstances. As everyone acts alone, the result is an erratic movement of the boat which rapidly undergoes a series of disasters until the weaker passengers and eventually all passengers have lost their fight. Alternatively, the passengers could begin negotiating a plan how to move jointly through the danger area, based on systematic acquisition of information about the environment, and use their paddles in a coordinated manner, with more efforts applied by the stronger passengers. This image is somewhat similar to the situation that mankind is facing with regard to climate change (Fig. 1). As a result of anthropogenic causes planet Earth is being driven into unknown regions of the climate system, posing one of the gravest threats facing mankind. As we know from climate history, the planet went through different climate extremes, from the ice ages to periods of extreme heat, with rapid changes sometimes occurring in a few deacades. A visible indicator was the sea level which fluctuated by dozens of meters. Every country, every industry and every citizen contributes to greenhouse gas (GHG) emissions, which like paddles drive us through the climate landscape. The question is whether and how fast we are learning to use these paddles in a coordinated manner, by developing goals and adaptive strategies to steer the planet through the landscape, based on limited but increasing information. At least there is agreement on the general goal. While global warming affects the integrity of natural and social systems, with severe risks to their stability and security, the UN Framework Convention on Climate Change (UNFCCC) demands stabilization of atmospheric greenhouse gas concentrations at levels that “prevent dangerous anthropogenic interference with the climate system”. With the entry into force of the Kyoto Protocol in 2005 the international community has established a first set of cooperative instruments to address the problem of global warming. A lack of agreement on the underlying causes, expected risks and required actions related to longterm climate change impedes further progress. An unprecedented degree of international action and cooperation is required to speed up emission reductions and technological change in the energy sector. Substantial progress is hampered by the expected costs of the transformation process and by partial interests that undermine the required cooperation. To overcome the hurdles associated with the “tragedy of the commons” an evaluation and negotiation process across all levels is needed, involving citizens, firms, institutions and states. Moving beyond Kyoto is a challenge for the policy process that is supposed to implement the longer-term objectives and manage the potentially severe implications in case of failure (Pershing/Philibert 2002, O’Neill/Oppenheimer 2002). It is also a challenge for the scientific community that increasingly becomes involved in value judgments and soft science issues that require innovative integrated approaches which support the policy process on different levels. 2

While the IPCC Third Assessment Report (TAR) paid some attention to these issues (IPCC 2001, WG3), a more systematic coverage is expected in the Fourth Assessment Report. Long-term climate change poses not only a challenge for the decision-making process but also for the decision methods and tools applied in this process (Sprinz 2005). Decisions under deep uncertainty and complexity hardly fulfill the requirements of established rational choice methods such as optimal control and game theory, lacking perfect foresight and complete information. Going beyond these approaches, adaptive decision-making and cooperative management allow to adjust actions and targets to the limited knowledge about the state of the climate system and the capabilities available to decisionmakers as well as the complex socio-economic interaction that undermines predictability.

Figure 1: Moving towards a dangerous climate landscape The Ultimate Objective: From Interpretation to Implementation The interpretation und implementation of the ultimate objective defined in Art. 2 UNFCCC is a key issue in climate negotiations beyond the first commitment period. Art. 2 calls for “stabilization of greenhouse gas concentrations in the atmosphere at a level that would prevent dangerous anthropogenic interference with the climate system. Such a level should be achieved within a time-frame sufficient to allow ecosystems to adapt naturally to climate change, to ensure that food production is not threatened and to enable economic development to proceed in a sustainable manner.” The three criteria – ecosystems adaptation, food security, sustainable economic development - help to specify and constrain the time-frame within which the 3

stabilization level should be achieved. Before addressing the implications, some of the key terms will be discussed here (for a more extensive study see Ott etal. 2004). Stabilization Level and Time-frame A key issue is to translate the ultimate objective into a stabilization level and time-frame that prevents dangerous interference and violation of the three conditions. „Stabilization“ is a process to reach an equilibrium of the climate system at which sources of GHG-gases and removal processes are balanced on average. (IPCC 1994, p.11). There is a range of possible future projections of carbon concentrations, resulting from the actions of all world citizens. The process is driven by decisions and negotiation across all levels, from local to global, taking into account the benefits, costs and risks of alternatives. Even if a prescribed stabilization path is principally technically possible, it may be hard to reach politically and economically. The scenarios outlined in IPCC (2001) indicate that the baseline emissions of CO2 would result in GHG concentrations ranging from 500 to 900 parts per million (ppm) until 2100, but stabilization would not yet be reached within the 21st century. Even for immediate and stringent emission reductions, past emissions have lead to nearly a doubling of pre-industrial CO2-concentrations. An increase in global mean temperature of more than 1oC seems unavoidable, as well as associated damages. The German Advisory Council on Global Change (WBGU 1998) proposed a tolerable magnitude of 2°C global temperature increase compared to the pre-industrial era, and a rate of temperature increase of 0.2 °C per decade, a view that is shared by many researchers and the European Union which has adopted the atmospheric concentration target 550 ppm. The challenge for decision-makers is to choose emission trajectories that are both feasible and represent reasonably ambitious levels of stabilization. A “safe level” has to be achieved within a time-frame which is compatible with the overall goal, somewhere between “as soon as possible” and “as late as necessary”: “Being too late should be avoided for environmental reasons, being too early for economic reasons.”(Ott etal. 2004) Due to the inertia of the climate subsystems and uncertainties it is difficult to determine dangerous points of no return which give sufficient reason for risk precaution. The costs of stabilizing CO2 concentrations in the atmosphere increase as the targeted stabilization levels decline. To meet the Kyoto targets costs range from about US$ 20/tC (tons carbon) up to US$ 600/tC without emissions trading, and from about US$ 15/tC up to US$ 150/tC with trading (Annex B countries). For several countries, GDP effects range from negligible to a several percent increase. The exact magnitude, scale, and scope of ancillary benefits and costs will vary with local geographical and baseline conditions. Danger Levels Human societies are sensitive to the effects of climate change which will affect human well being, income distribution, and adaptability to climate change. Vulnerable systems include water resources, agriculture, forestry, human health, human settlements, energy systems, industry, and financial services. The proposition „prevent dangerous anthropogenic interference“ assumes that mankind is able to avoid inacceptable danger. This implies that is possible to find thresholds separating regions that are too dangerous and others that are not. At present, there is no common understanding on the long-term goals and the criteria to evaluate dangerous interference. The term „dangerous“ is inherently related to normative questions and cannot be reduced to a strict

4

scientific meaning. IPCC 2001 has identified possible categories of serious concerns: unique and threatened ecosystems; distributional impacts (justice); aggregate impacts (side-effects); extreme climate effects and large-scale singularities. If GHG concentrations cannot be stabilized at tolerable levels, mankind may face potentially disastrous consequences. Dwindling glaciers, changing ocean currents and precipitation patterns, sea-level rise, harvest losses, degradation of biodiversity, as well as indirect impacts such as hunger, poverty and environmental refugees, can effect millions of people. Droughts and floods, heavy storms, forest fires and other disasters can be devastating on a short time scale (Hare 2006). Large-scale climate change can induce grave social and economic disturbances and instabilities in different world regions which could generate or intensify social instability, conflict and human insecurity on multiple levels (Barnett (2003), Scheffran/Battaglini 2007). The dangers will be geographically dispersed and depend on the vulnerability and adaptation potential of regions and actors. In addition, the potential for abrupt climate change with potentially catastrophic consequences cannot be ruled out (Keller et al. 2006, Alley et al., 2003, Stocker 1999). The loss of unique coral reef-systems is likely already for a moderate temperature change. An uncertain and potentially threatening event is the disintegration of the West Antarctic Ice sheet (WAIS), with massive sea-level changes by 4-6 meters (Oppenheimer and Alley 2005). A shutdown of the North Atlantic Thermohaline circulation (THC) would have far reaching, adverse ecological and agricultural consequences (Schlesinger et al. 2006, Rahmstorf and Zickfeld 2005). Positive feedback from warming may trigger the release of carbon or methane from the terrestrial biosphere Equity and Normative Issues Equity is not explicitly mentioned in Art.2 although it is important to understand who is potentially affected by climate change and how the risks are regionally distributed. Some countries are more vulnerable than others due to their natural geographic and socio-economic conditions and the lack of adaptation capabilities. First of all, marginalized and vulnerable communities are threatened, more than well established and stable communities, although wealthy countries are not immune to social instability. Developing countries, which are less responsible for global warming, would be affected much stronger and would be less capable to take countermeasures. Given the asymmetries, it is a challenge to find a fair allocation of emission rights respecting these limits and to counter balance risk extremes in certain regions. Industrialized countries with high per-capita emissions agreed in the Kyoto Protocol to cut them down and to establish cooperative instruments such as Emissions Trading, Joint Implementation and the Clean Development Mechanism to facilitate emission reductions. An associated issue is to find mechanisms to allocate emission limits and permits from global levels to regional, national and local levels, including individual firms and consumers. Although Art.2 was adopted in consensus by the negotiating parties, its specification is hampered by conflicting interpretations. While concrete targets beyond the Kyoto Protocol have not yet found agreement, increasing attention is being paid to the interpretation and implementation of Art. 2 (e.g. Schellnhuber et al. 2006). To specify “danger standards” should be a common interest that is compatible with the negotiation and decision competence of the Parties of the UNFCCC. Building on universal ethical norms as guidelines for orientation and argumentation helps to build legitimacy and acceptability of negotiation results. Determining a level of stabilization can partly rely on scientific observations or measurements but is not only a scientific task. Tolerable 5

danger levels are set by decision-makers based on a combination of interests and reasonable judgement, using the state-of- the-art in climate science and ethical criteria as inputs. Decisionmakers will have to operationalize the question which scale of regional and temporal disruptions are acceptable to them and how to bridge unequally distributed climate impacts which may be positive in some regions and negative in others. Lack of full scientific certainty is no sufficient reason to postpone precautionary measures “to anticipate, prevent or minimize the causes of climate change and mitigate its adverse effects” (as required by Art. 3.3 UNFCCC), in particular if they are irreversible and cannot be compensated for. Thus, a reasonable and systematic specification of the overall objective with regard to key parameters (stabilization level and timeframe, regional impacts and danger levels, ecological, food and economic criteria) is still needed to operationalize and implement the Convention. Specifying the Three Criteria The ultimate goal of stabilization of GHG concentrations at tolerable levels is augmented by three criteria: adaptation of ecosystems, secure food production, sustainable economic development. These can be used as constraints which must be satisfied during the process of reaching a stabilization level. Specifying these criteria, including potential conflicts and tradeoffs, leads to important conditions for implementing the UNFCCC’s ultimate goal (for a more comprehensive assessment see Ott et al. 2004). Adaptation of Ecosystems Many natural systems are vulnerable to climate change and have limited adaptive capacity, such as glaciers, coral reefs, mangroves, arctic and mountainous ecosystems, wetlands as well as biodiversity hot spots, among others. Some of these systems may undergo significant and irreversible damage. The laws governing the pressure on ecosystems and their adaptation are highly complex and not yet well understood. A crucial issue is whether ecosystems remain intact despite external influences. Adaptive ecosystems are able to preserve the essential qualities that define their identity and existence through feedback cycles that maintain stability within viable limits. If these feedback cycles are disturbed as a result of human interference, rapid, even catastrophic changes may occur, leading to the loss of valuable functions of the ecosystem. Whether these functions can be preserved against likely disturbances indicates the stability and resilience of the ecosystem. The closer the ecosystem comes to its viability limits, the more restricted is the admissible range of actions. Global average temperature change needs to be compatible with the survival of ecosystems which translates into maximum allowable emissions. The time-frame for ecosystem adaptation determines the admissible speed of climate change. In many ecosystems climate change results in a slow change of external factors which may have no directly observable effects as long as critical thresholds are not reached. Thus, over a short time horizon, these ecosystems would be able to adapt to the changes in their environment. With a longer time horizon, such ecosystems could move into a state of less stability where they are threatened to become destroyed or irreversibly damaged. Thus, ecosystems exhibit adaptive capacities up to a certain threshold beyond which they break down. These thresholds are often difficult to determine.

6

Adaptation capacities for ecosystems may accept some local disruption but no large scale disruption. Many local to regional ecosystems are already drastically changing and some have adapted to climate change by moving to a different location, by modifying their internal flows, or by replacing the structure of ecosystems. For instance, coral reefs respond to increasing water temperatures. If ecosystem change cannot be avoided, decisions need to be made as to which parts of an ecosystem are to be preserved. This requires an assessment of the value of ecosystems to society balanced against the cost of climate policies. Generally, managed ecosystems are better able to adapt to climate change, and ecosystems with poor resource endowments are more vulnerable (IPCC 2001, WGII). In addition, managed ecosystems in poor economies have a limited capacity to adapt and are more exposed to climate change. Thus, the benefits of preservation and the costs of adaptation or mitigation are distributed unevenly between rich and poor economies. Securing Food Production Agriculture is exposed to stochastic weather events which are driven by the uncertainties of longterm climate change. While highly productive intensive agriculture is rather fragile, it can be stabilized through management practices adapting to external changes. Thus, managed ecosystems which provide most of the food are less vulnerable to climate change than natural ecosystems or subsistence agriculture. The conversion of reserves into land for food production may run into conflict with the objective of ecosystem preservation. Potential negative impacts on crop-productivity are compensated by other adaptive reactions such as changes in technologies or crop patterns. More efficient production technologies, modified crops, and optimised inputs have continuously reduced the area of land to feed one person by a factor of ten compared to Malthus’s time. Despite the fact that still a substantial number of people suffer from hunger, it is estimated that 8-10 billion people can be fed with today’s technologies, even under unfavourable climatic conditions (Tilman et al. 2002). Therefore, in the next decades, climate change is not expected to depress global food availability, but it may increase the dependence of developing countries on food imports and contribute to food insecurity for the most vulnerable groups and countries. The main problems are regional imbalances and insufficient purchasing power of people in poor regions. A move to secure local food production might require a more restrictive climate policy with potentially high negative impacts on incomes. The interaction of future development in soil fertility, water availability, advanced crop designs and land availability is complex and will be even more difficult for regional predictions or scenarios. Although improved crop varieties provide higher yields – often together with an expansion of other inputs such as fertilizers or pesticides – they also become more susceptible to adverse impacts such as varying climatic conditions or diseases. Sustainable Economic Growth Presuming that restrictions on GHG emissions will slow down economic growth (which may not always be the case), the requirement of not interfering with sustainable economic development

7

implies that climate policies should achieve a desired emission path at lowest costs. The costs of preventive measures basically include the GDP reduction due to a reduced input of fossil energy sources, and the investments, consumption losses, user and adjustment costs on the way to new energy systems. Sustainable economic development is further affected by the vulnerability of economic systems and the damaging impact of climate change on economic growth. There are considerable uncertainties with respect to several economic factors, such as the rate of innovation, capital accumulation, and institutional reform. Under these circumstances predictions about long-term economic development become almost impossible, not to speak of surprises such as civil wars and natural disasters. There is little known about institutional changes of societies which influence economic growth and human welfare, such as legal protection, property rights, functioning markets, infrastructure, etc. Most of the very poor countries suffer not only from a lack of resources and human capital, but also from a lack of institutions which support and enable economic activity. A regional focus could lead to mitigation options where for some regions a slower climate change can be more desirable than for others. Since the greenhouse effect is a global phenomenon requiring coordinated action, these differences need to be settled in climate negotiations. Some of the Small Island States are existentially threatened through sea-level-rise and extreme weather events. For them dangerous climate change may be impossible to prevent, even if strong actions were taken worldwide. In addition, adaptive measures against climate change impacts may be so costly that they could be detrimental to sustained economic development. Consequently, such states will inevitably need to rely on international support. A focus on the world economy rather than a regional or national focus on growth would require balancing actions in the international community to compensate for expected regional imbalances, beyond the commitments of Annex B countries in the Kyoto-Protocol. Climate policies exert different impacts on economic growth with a time horizon of decades, depending on the time path over which fossil fuels will be replaced through alternative energy sources or more efficient energy uses. Since climate change impacts are not expected to significantly slow down economic growth for a period of 10 to 20 years, the initial impact on economic growth will predominantly come from preventive measures. Economically inefficient policies would result in a slower path towards a target level of GHG concentrations. The faster the replacement of fossil energy by non-fossil sources, the higher the adjustment costs because existing capital stocks with high energy intensities need to be depreciated faster than planned. On the other hand, the longer the introduction of non-fossil energy supplies is postponed, the more capital is directed towards fossil energy facilities which usually have a long life-time of several decades. The basic trade-off between shorter and longer time horizons is described in Ott etal. (2004):80: “Policies which impose little constraints on short term economic growth coincide with higher emissions and an increasing scarcity of fossil energy sources in the longer run. Policies which start mitigation early and to a significant degree will slow down economic growth in the short-run improve the growth potential in the long-run by preserving natural resources including fossil sources and by reducing the negative impacts of climate change.” While policies with a very long-term horizon may be desirable, attempts to influence economic development over half a century or more do not rest on solid grounds because of today’s limited knowledge and influence. More realistic is a focus on time scales of one or two decades. 8

Principles of intergenerational justice and fair distribution of benefits and costs of climate policy measures demand a more disaggregate regional and sectoral focus, balancing short-, medium-and long-term sustainability requirements. Conflicts and Trade-offs between the Three Criteria To meet the three criteria of ecosystem adaption, food security, and sustainable economic development it is essential to discuss their interdependencies and trade-offs. Using measurable concepts allows to ensure compliance as a prerequisite to operationalize Art.2 for policy making. Some difficulties emerge from the fact that these criteria are influenced not just by climate change, but also by a variety of other factors, including food markets, population growth, economic development, technical change, land use patterns, etc. Furthermore, actions that are adequate for one constraint may adversely affect one of the other two. In the following some of the trade-offs and potential conflicts will be discussed pairwise.  The stability of ecosystems and food security both depend on the environmental space available and the degree of energy flows and materials. Natural ecosystem adaptation to climate change often requires space and time to migrate to locations where species can cope with the new environmental conditions. Producing food depends on the area available for agriculture and on the intensity of land use, both of which conflict with the needs of ecosystems adaptation. Thus, the trade-off between ecosystem adaptation and food production becomes important in regions where natural areas and agriculture compete for scarce land. In industrialized countries where the agricultural sector is capital intensive and uses high-yield crops, there is little space for biodiversity and the preservation of natural systems. On the other hand, higher productivity leaves more room to preserve significant areas from intensive agricultural use, thus giving more space to natural ecosystems. Many less developed countries face a shortage of arable land, leaving only little or no area for natural ecosystems. Ecosystems in tropical and subtropical zones tend to be more vulnerable than in temperate zones although agricultural practices often allow for a larger biodiversity and less stress to natural processes than in temperate zones.  Food security and sustainable economic development do not seem to be in an apparent conflict, as the historical experience suggests. High agricultural productivity and sufficient food supplies are highly correlated with per-capita incomes and advanced technological knowledge, i.e. rich and technically sophisticated economies also produce sufficient food, often too much food as the EU case shows. Productivity growth in agriculture has even outpaced that in industry in industrialized countries. High incomes also tend to create sufficient and effective demand with sufficient price incentives for a modern agricultural sector. However, in low income countries with little demand and low productivity in agriculture regionally disaggregated strategies for food supply are required to overcome the vicious circle of insufficient price incentives and low agricultural productivity. Most appropriate is a balanced growth policy to stabilize traditional agriculture through programs supporting small farmers and to provide savings for industry development. Conflicts might arise in the future related to institutions which are insufficiently able to strike a balance between allocating resources towards the agricultural sector and towards industrial development.

9

 Ecosystems adaptation and extended economic development that only considers economic and social criteria are in potential conflict. Economic growth relies on the use of natural resources, including space, energy, raw materials, or nature as a sink for emissions from production and consumption processes. Ecosystems may undergo stress by economic growth (whether induced by climate change or not), reducing their natural ability to adapt to climate change. These relationships have been intensively debated with regard to the “Environmental Kuznets Curve”, the supposedly inversely u-shaped relationship between economic development and environmental degradation. While in many cases there is weak evidence for this hypothesis, this has not been confirmed for CO2. Thus, economies with rising incomes will not automatically start reducing their GHG-emissions which in turn will not lead to a decoupling between economic growth and ecosystem stability with respect to climate change. A moderate GHG stabilization level at 500 ppm CO2-equivalents and a short time frame would allow a natural adaptation of ecosystems. This would require a fast reduction of GHG emissions and place a considerable burden on economic development, including drastic price increases for fossil fuels and large investments in a new non-fossil capital stock. On the other hand, the choice of a very long time-frame for stabilization could involve an extended process of increasing emissions before a turn is made towards reductions. It has been argued that this process would buy time to develop low-cost non-fossil energy sources and get a better understanding of the climate system, accepting the risks from climate change and ecosystem degradation for some time. Most preferable is a win-win situation in which strategies and actions support sustainability in both natural and social systems. Integrated Assessment and Adaptive Decision-making Optimal and Adaptive Control of Climate Change Integrated Assessment combines the dynamic interaction between natural and socio-economic systems to understand the implications of decision-making on future climate change. Emission scenarios use computer simulation to project carbon emissions, concentrations and temperature change into the future, based on plausible parameter variations. Forward approaches determine a set of emission trajectories from initial conditions by variation of scenario-dependent parameters, while inverse approaches calculate admissible “funnels” of emission trajectories compatible with a tolerable or targeted temperature range. Comparing and combining these approaches provides a framework for long-term climate decision-making to avoid critical thresholds that trigger abrupt change. Optimal control methods seek to maximize time-discounted utility functions (welfare), iwhich include expected benefits, potential climate damages and the costs invested. The Dynamic Integrated assessment model of Climate and the Economy (DICE) and its regional variant RICE have been used in many previous studies to design optimal climate policies (Nordhaus 1993, Nordhaus and Boyer 2000). The approach is based on a globally aggregated optimal growth model, with a Cobb-Douglas production function that describes the flow of economic output, depending on capital, labor and technology. The merits and limits of utility optimization have been well recognized. Global utility functions are based on the assumption of a world decisionmaker who has complete knowledge and selects an optimal time-discounted control path (for a critical analysis with regard to integrated assessment of climate change see Füssel 2006).

10

An alternative is the Tolerable Windows Approach (TWA) and similar concepts (safe landing, guardrails) that restrain and adjust the path of GHG emissions to keep global average temperature change within viable bounds of natural and social systems and avoid critical levels of danger, in particular disastrous events.(Bruckner et al. 1999, Petschel-Held et al. 1999).1 The admissible corridor defined by guardrails can be perceived as the space to maneuver for future climate policy. The admissible domains of the climate system and the emission paths within the boundaries are given by assessments and judgments of climate change, taking into account vulnerabilities and adaptive capacities, as well as critical thresholds for phase transitions and extreme events that may cause qualitative system changes (Schneider 2004). The task is to identify key control variables and regulation mechanisms for decision-making to maintain the admissible domain in order to avoid “intolerable dangers” of climate change, whoever defines the intolerable level. An extension of optimal control and TWA is adaptive control where actions are taken according to rules that respond to the actual state of a system in a prescribed direction. Adaptive control approaches constrain and adjust the path of GHG emissions towards a target or to stay within a viability domain of the climate system (e.g. a certain range of carbon concentration or globalaverage temperature change). Actors decide and act on the basis of incomplete knowledge, usually restrained to a spatial and temporal window of attention. Within this window, actors sequentially select actions according to decision criteria, e.g. to maximize a utility function, to pursue any other target or to stay away from a dangerous area. One possible rule in climate policy could be: when future projected emissions exceed a critical temperature threshold actors increase their investment into emission reductions until the projected path stays within the limit. This requires a definition of a critical temperature threshold and knowledge about its distance from the current position as well as the rate of temperature change, based on observations and models describing the interaction between carbon emissions and temperature change. Adaptive rule-based approaches are adequate in complex and deeply uncertain situations where actors do not know enough about the future to calculate long-term optimization. While the data and knowledge of the dynamics are bound by uncertainty, a forward looking actor can observe and project a future channel within which the dynamics is likely to stay. If this channel misses the target or hits a “forbidden” area, actions are required within the available capacity of resources. Continuously updated scientific information is essential to estimate whether the combination of current state and future trend is tolerable within given limits. If not, some sort of “speed control” is required, similar to pushing a brake or accelerating in a foggy environment where speed is adapted to sight, destination and unexpected events. Such an adaptive control strategy can be implemented via decision rules and response functions, taking into account actions taken by other actors. A crucial issue are the timescales. There can be a considerable time lag between emission reductions and their impact on the climate system. Temperature effects may be expected 20 to 50 years after peak emissions of CO2 whereas sea level changes may occur hundreds of years after concentrations have stabilized. This problem is aggravated by the fact that due to inertia of the socio-economic system the effect of policies will be delayed, too. This concerns in particular the 1

The mathematical context for TWA is viability theory which applies regulators to keep a dynamic system within viable constraints (Aubin/Saint-Pierre 2004). For an extension towards probabilistic guardrails see Kleinen 2005.

11

replacement of infrastructure and technology, such as buildings, power stations or transport systems, which can take several decades or even more. As a consequence, considerations of time lags seem to be essential for adequate political decisions. Time discounting represents the degree to which decisionmakers take the future time horizon into consideration. How far this horizon reaches into the future or into the geographic environment depends also on ethical criteria and precautionary principles that can shape control strategies. Discounting provides a quantitative measure for comparing decisions with consequences occurring at different times. Because of the uncertainties of the distant future, choosing an appropriate discount rate for impacts spread out over decades or even centuries is a controversial issue. The prescriptive approach in climate analysis favors substantial investments for climate mitigation policies, corresponding to low discount rates. The descriptive approach favors spending on immediate social needs, justifying high discount rates. This “wait and see” policy is consistent with market realities and short-term investment priorities of international funding institutions. Uncertainty, Probability and Risk While empirical evidence of climate change is mounting, significant uncertainties still remain (Murphy et al. 2004, Stainforth et al. 2005). Major causes of uncertainty are: precipitation patterns which determine the regional distribution of severe impacts; the capacity of the biosphere and oceans to remove CO2 from the atmosphere; the physiological reaction of plants on increasing CO2 concentrations; and the regional impacts of climate change on ecological and social systems. Actors do not exactly know the system state, the impact of their actions and also the values themselves. At present, a considerable part of the U.S. effort is aimed at reducing the uncertainty concerning climate change impacts, rather than reducing the impacts themself. Uncertainties are an obstacle to reaching agreement on a particular concentration level, but they do not justify a delay of necessary action. Dealing with the problem, uncertainty analysis of future global temperature change became a major research topic in recent years (Wigley/Raper 2001; Forest et al. 2002; Allen et al. 2000; Dessai/Hulme 2003; Giorgi/Francisco 2000). One approach would be to define goals for different time-periods sequentially, whereupon later goals are made dependent on the achievement of goals in earlier periods (hedging strategy) (Yohe et al., 2004). To test for the relevance of uncertainty, probability distributions are required, calibrated against data of key factors. Most significant are climate sensitivity (Andronova/Schlesinger 2001), the carbon intensity of energy systems, and the evolution of population growth and economic production. The world is facing deep uncertainty in our understanding of climate thresholds when systems models and probability functions are unknown (Alley et al., 2003; Lempert, 2002). Anthropogenic greenhouse-gas emissions increase the likelihood of crossing critical thresholds, possibly leading to abrupt events. For instance, the potential THC breakdown can lead to drastic changes in danger levels and reactions, and thus should be avoided with uncertainty-dependent safety margins. Risk assessment and management provide important tools to deal with the problems. Bayesian uncertainty analysis of the probability distribution of variables is an important instrument to estimate the likelihood of certain outcomes from prior knowledge. Novel

12

approaches such as belief functions (Kriegler 2005, Kriegler/Held/Bruckner 2006) define lower and upper bounds for the probability. Quantification of uncertainty is more challenging than the more common approach of building a set of self-consistent scenarios without a systematic analysis of how likely the various scenarios are. Rethinaraj 2005 and Singer etal 2006 demonstrate that it is mathematically and computationally tractable to develop adequate theory-based models of global climate change and global average temperature change and calibrate their a priori uncertainty distributions against appropriate time-series data. Probabilistic and adaptive approaches can be combined. Since the observed state and its change have a range of uncertainty, the future path can be predicted with a probability distribution. Adaptation would adjust to uncertainty by keeping a safety distance for the lower tail of the PDF (worst case perceptions) or be more relaxed by accepting a higher threshold for the higher tail (best case assumptions). In this context a thorough and quantitative uncertainty analysis is an important contribution to decision-making and risk assessment on longterm climate change. Risk perception would take into account both the probability and damage of critical events which depend on the stability and complexity of the affected systems and can be influenced by mitigation and adaptation strategies. Modeling of the coupled climate-economy interaction, using the best available data, contributes to a reduction of risk perception (Fig. 2).

Figure 2: Impact of mitigation and adaptation on risk factors within a modeling framework

Mitigation and Adaptation in Energy Transitions Decision-making on future energy paths can make significant contributions to preventing dangerous climate change. The task is to find the proper energy mix balancing economic and environmental aspects, taking into account technical factors, such as the cost per energy unit, energy productivity and conversion efficiency. An integrated evaluation seeks to balance welfare optimization, risk miminization and cost efficiency. Decision makers have to decide which mix 13

of mitigation and adaptation to pursue. Mitigation makes sense before an event occurs to prevent it, while adaptation after it has occurred. Mitigation has largely global effects, whereas the benefits of adaptation usually apply to those who invested into such policies. Practical measures to reduce emissions include behavioral and technical changes. Energy savings and efficiency improvements contain a considerable potential to reduce demand; energy technologies with less carbon emissions per energy unit reduce environmental impact. Technological options are particularly relevant to infrastructure awaiting retirement in the near future. A critical question is under which conditions endogenous technical change would support a transition towards low-emission technology (Edenhofer et al. 2006). Producers and consumers tend to favor low-energy and low-emission technologies if they provide higher benefits and lower costs. A crucial issue is to determine the threshold costs for switching which could be used to design policies that reduce these threshold costs, legal regulations, taxes, emissions trading, subsidies and other more conceptual approaches that support cooperation and coalition formation in favor of a sustainable transition. A typical scenario based on optimal economic growth and slow technical change leads to a doubling of human population by the middle of this century, an increase of GDP per capita and energy per capita, a doubling of carbon emissions and atmospheric carbon as well as a temperature rise of almost 3oC which results in considerable climate damages. When climate damage becomes significant, investment into low-emission technology begins to accelerate which speeds up the transition towards decarbonization and a reduction of carbon intensity. Because of the initial delay, a large temperature change and the severe climate damages at the end of the century cannot be avoided (Scheffran 2006c). The outcome considerably differs if adaptive target setting is used to achieve a temperature limit of 2oC by end of the century. Translating the temperature limit into required emission reductions, results in an earlier and more substantial transition of investment into low-emission technologies. Compared to the first scenario, GHG emissions would be reduced by about 50% and atmospheric carbon would stabilize at about 450 ppm (Scheffran 2006c). Carbon intensity is reduced more rapidly, with much lower amount of total carbon emitted and associated damages, at the cost of income losses. Multi-agent Interaction in Climate Policy Multiple actors are shaping the interaction between the climate and the economic system as they can choose targets as well as actions. Decision-making is complicated by the number of actors and multiple levels involved that interfer with each other. At global levels of decision-making the main actors are usually governments of nation states or groupings among them, often clustered along regional boundaries. At local levels individual citizens are key players who affect or are affected by global warming. The multi-level process between local and global decision-making passes through several layers of aggregation (from billions of citizens to a few diplomats representing their countries), with each layer having its own decision procedures for setting targets and implementing them into real actions. The outcomes for each actor are highly dependent on the actions of other actors. Given these complexities, a crucial issue is how the world can act together and cooperate on climate change, managing the transition from individual competition to cooperative action.

14

Multi-actor and multi-level decision-making can follow a top-down approach where global decision-making bodies define global targets for emission reductions based on scientific assessment and an evaluation of which degree of dangerous climate change is tolerable. The task is to implement these target at lower, in particular national levels. In a bottom-up approach local actors such as citizens, consumers and companies pursue their individual interests, having an impact on higher levels, e.g. by electing municipal and national governments or by selecting products with more or less environmental impacts. In reality both approaches interfer with each other at each level and could potentially lead to conflict (see Fig. 3).

Figure 3: Multi-level decision-making on climate targets and actions

The collective action problem is to agree on an emission path that avoids dangerous climate change and to make sure that the cumulative emissions by all human beings will not exceed this limit. Assuming that there is an agreed cap on aggregate emissions, it is then a challenge to find institutional mechanisms to ensure that individual limits are assigned to each actor and that their compliance is ensured, avoiding the tragedy of the commons. Despite diverging preferences, some cooperation is indispensable. To describe, understand and model the complex micro-macro links and the interaction of conflicting interests across all levels is a challenge for projecting future GHG emissions. Various tools have been developed to understand the interaction among multiple actors in climate policy and to help identifying possible solutions (Scheffran 2006b). Game theory provides a framework for analyzing interdependent decision-making and negotiations on climate change.(Carraro/Filar 1995, Svirezhev et al. 1999, Finus 2001, Carbone et al. 2003, Kemfert 2004, Haurie/Viguier 2005). If actors are unable to improve their situation 15

to their own favor they are in a so-called Nash equilibrium. If actors may improve only at the cost of other actors this corresponds to a Pareto optimum (collective equilibrium). In a dynamic (repeated) game situation the actors mutually adapt their targets, values and actions to those of other actors to change the situation to their own favor (Scheffran 2002ab, Scheffran/Pickl 2000, Krabs/Pickl 2003). With an increasing number of actors who follow given response patterns agent-based modeling becomes more appropriate (Weber/Barth/Hasselmann 2003, Weber 2004, Billari et al. 2006). For multiple criteria a conflict may occur if some criteria lead to positive, others to negative evaluation. The conflict can be diminished by pursuing actions that improve all criteria (win-win). Alternatively, emphasis will be given only to the most-relevant criteria, by priorising or lexicographic ordering. Cooperative approaches include the international transfer of investments and technologies to shift the composition and learning rates of the energy system towards emission reductions. In negotiations actors adapt and restrain their freedom of action to achieve mutual benefits, reduce costs or diminish risks. To implement Art.2, states need binding and verifiable agreements to avoid a prisoners’ dilemma. Actors can adapt their climate targets or merge their resources and investments in coalitions to be better off by acting together rather than acting alone. (Eisenack et al. 2006). Coalitions are reasonable if individual action is insufficient or inefficient, e.g. in acquiring a critical number of votes or a critical mass of investment to realize projects. For Art.2 a critical mass of emission reductions is needed to prevent dangerous interference. In value-based coalitions actors seek agreement by adapting their positions and values to each other, as is the case with the Kyoto targets (Grundig et al. 2001). In resource-based coalitions actors merge their investments to generate joint benefits which are then distributed to the individual actors (Scheffran 2006a). An example are the Kyoto instruments. Coalition formation describes the transition from individual to collective action as a bargaining process where the probability of joining a coalition increases with the values actors expect from it (see Göbeler/Scheffran 2003). Coalition formation is more likely among actors with similar positions. For instance, countries tend to cooperate more closely if they prefer similar levels of carbon concentration or temperature change, as can be seen in the formation of country groups in the UNFCCC and Kyoto process. Market mechanisms are assumed to provide an efficient and cost-effective allocation between regions and businesses. Emission trading is designed to achieve emission reductions in regions and business sectors where they are least costly (for an adaptive dynamic model with stylized data for 11 world regions see Scheffran 2004, Scheffran /Leimbach 2006). From Conflict to Cooperative Management Actors, Positions and Conflicts in Climate Policy The 1992 UNFCCC represents a compromise between a wide range of different interests. The formula of “common but different responsibilities” assigned different roles for industrialized (ICs) and developing countries (DCs) in climate policy which led to a grouping of the UNFCCC member states into Annex I, Annex II and non-Annex I countries, with no commitments for the latter. Since the Kyoto Protocol was signed, the conflict on the distribution of different obligations became apparent. While the biggest DCs (China, India) turn into major emitters, the reasons why they refuse obligations to reduce their emissions are understandable, first of all their demand for equity in terms of development chances. On the other hand, the United States insists

16

that the commitments for Annex B countries be extended at least to the major DCs. With this conflict unresolved, future global climate policy may remain in a prisoner's dilemma situation. To overcome diverging interests it is important to build coalitions among those with mutual interests. To specify the conditions for a cooperative solution the trade-offs between growth and emission targets and the relation between DCs and ICs have been analyzed in Ipsen et al. 2001, with the example of the power-generation sector in China. Cooperation becomes feasible if the DCs voluntary participation in climate policy is in their own interests, e.g. because it supports local environmental goals or access to advanced technology. The prisoner's dilemma can be overcome if ICs and DCs agree on mutually beneficial cooperative solutions with regard to technology transfer and low-carbon investments to fulfill both global environmental and economic goals. The benefits of cooperative action need to be large enough to induce sovereign states to accomodate opposing local economic interests and resolve conflicts. Whether conflict or cooperation prevails depends on the positioning of key actors on targets and actions. On the European level there is a comparatively ambitious positioning of relevant institutions regarding emission reductions and concrete GHG levels, occasionally justified by reference to Art. 2. On the global level, while there is still a lack of positioning, the issue is already shaping the current negotiation process and contributes to a clustering of key players into potential coalitions. Those who want to push the agenda towards the stabilization goal (such as the member states of the EU) are facing strong resistance by those who want to refuse or postpone any commitments (such as the US or key countries of the G77). Countries such as Russia and some members of the Umbrella Group could be key players in setting the future agenda in one or the other direction. The heterogeneity of G77 induces potential conflicts within this group that may contribute to slowing further progress on achieving the ultimate objective. With increasing attention of developing countries to their own vulnerability to climate change the need for speeding up the process may prevail. To have support from G77, a “fair” allocation of emission rights seems unavoidable, based on equity principles such as equal per capita emissions. The factual diversity of corresponding proposals towards realization of equity might complicate this effort to some extent. Conflicting Positions on Art. 2 - Results from an Expert Survey In order to extend the empirical basis on the positioning of actors in international climate negotiations with regard to Art. 2 UNFCCC, and to identify potential conflicts and negotiation strategies, the author had the opportunity to perform interviews and distribute a survey among experts at the 8th Conference of the Parties of the UNFCCC (COP-8) in New Delhi, October 23 to November 1, 2002. COP-8 took place at a critical juncture between short-term commitments and longer-term obligations, marking a new phase of negotiations. While different viewpoints on the ultimate objective were exchanged, it became clear that developing countries would not give up their “right” for increasing emissions without serious concessions in other fields of the development agenda that satisfy the demand for global equity and poverty reduction. The survey was to acquire feedback from selected experts who were asked to evaluate their own position for a set of questions and to assign a position to each of four major players in climate 17

negotiations (United States, European Union, Russia, G-77 plus China), based on personal judgement. Due to constraints at COP-8, the survey could neither be comprehensive nor representative. Since half of the experts were researchers and about half were from Europe, the survey disproportionately represented the view of European researchers, which should be kept in mind when interpreting the data. The survey raised questions about the benefits and costs of the ultimate objective, the clarity of its meaning and its consistency, in particular the compatibility of the three associated criteria, the inclusion of equity considerations, the preferred stabilization level, and the years required for starting and concluding implementation of Art.2 (the time frame). Except for the latter numbers, experts were asked to assign a number between –5 and +5 to determine the ordering of positions. By responding to the questions, the experts determined a position for themselves and for other actors in a multi-criteria space. The results of the survey refer largely to the average values (m) over all experts and the respective standard variation (sv) for the diversity of viewpoints. These results may indicate specific opportunities and difficulties for constructive future negotiations on Art. 2. The most relevant results are described below (for more results see Ott et al. 2004).  Ordering of actors: It is striking that for most variables a certain order of positions is assigned. Compared to an “optimistic” hypothetical actor who associates Art. 2 with high benefits, low and acceptable costs, high clarity and compatibility of the three criteria, as well as high relevance of equity, early implementation, low stabilization level and short time-frame, the experts put themselves in the first place, followed by positions ascribed to the EU, Russia, G77/China, and finally the USA at the other (more “pessimistic”) end of the spectrum. The multicriteria chart of Fig. 4 visualizes the differences between actor positions for key variables and the potential for conflicts and coalitions. A set of positions near the periphery represents more optimistic actors (the experts and the EU), while the position set in the inner core of negative variables represents actors critical to Art. 2 (USA).  Potential fields of conflict: Among all variables, the equity issue shows the largest diversity of opinions, indicating a major point of conflict in future climate negotiations. A second field of conflict is the perceived incompatibility between economic and ecological conditions, not so much because of the diversity of positions but rather due to the fact that - on average - all actors (including the EU) are found in the negative range. This implies that some inherent conflict is expected between the condition of ecosystem adaptation and enabling of sustainable economic development. The fact that this variable receives the lowest score for all actors means that the experts see a still unresolved issue. There is a wide range of views about benefits and costs which could make agreements to become more difficult to achieve. Rather high average values and a smaller range of opinions occur on the compatibility of ecology-food and economy-food which implies that the experts see here less diverging positions.  Implementation, stabilization and timeframe: Negotiating the implementation of Art. 2 was preferred by the experts themselves to start around 2005, which is about two years earlier than for the EU. Other average implementation years are 2011 for Russia, 2015 for G77/China and 2017 for the USA which also shows the maximum variation of 12 years. Experts prefer an average stabilization level of 482 ppm CO2-equivalent, for the EU they assign an average level of 533 ppm. On the opposite end is the USA with an average of 724 ppm. The average position of G77/China is rather homogenous at 600 ppm, while positions ascribed to Russia are quite

18

heterogeneous (with an average of 626 ppm). The experts’ valuations are variable with regard to the required time-frame for stabilization. The averages are quite similar for the experts’ own position (2066) and that of the EU (2074); Russia (m = 2093) and G77/China (m = 2086) are supposed to prefer longer time-frames and show a similar distribution. The USA is an outlier with the average time-frame expanding to 2113 (Fig. 5).

Figure 4: Multi-criteria chart of average positions on Art. 2 issues for different actors

Figure 5: Average positions of actors on stabilization issues. 19

 Potential coalitions: The clustering of data provides some indication about potential coalitions. Not surprisingly, with regard to most variables, there is a rather high coincidence between the EU position and the experts’ own viewpoints (in particular of researchers and NGOs). This coincidence which can be observed in climate negotiations is largely independent of the origins of the experts. For net benefits of Art. 2, a grand coalition appears to consist of EU, Russia, G77/China, in line with the experts’ self-assessments. Only the USA is found in the negative benefits range. With regard to the timeframe, nearly all country groups are assumed to set stabilization targets before the year 2100, except the USA following with more than a decade delay. The experts’ preference for equity is similar to G77/China and comes close enough to the EU to support a coalition. But the gap between the experts and the EU is here more significant than for any other variable. The issue of the clarity index suggests that no actor appears to have full clarity of the meaning of Art. 2. The experts themselves and the EU achieve a value near +2, whereas the other actors plot near zero (USA) or reach even negative values (Russia, G77/China).  Uncertainty of positions: It is interesting to note that standard variation (sv) is largest for the group of interviewed experts themselves for most variables while viewpoints about others are more constrained (see Figs. 6 and 7). This sounds reasonable since viewpoints are involved with quite heterogeneous backgrounds. Remarkable are the following exceptions: the lowest variation is achieved for “benefits” and “costs” which indicates that the experts basically agree on the significant benefits and low costs from Art. 2. The standard variation is also low for “equity” which implies that there is little uncertainty about its significance. For most variables Russia receives the lowest of all uncertainties, while nearly all variables on the USA exhibit the highest variation, immediately followed by G77/China. This result seems unexpected because the USA is often treated as a candidate with a clear position, while G77/China is supposed to have a wide range of positions. A similar tendency can be observed on “implementation”, “stabilization level” and “time-frame” (Fig. 7). “Implementation” shows a steady increase in uncertainty from experts to USA. Nevertheless, “stabilization level” shows a boost in variation of ascribed positions for Russia, G77/China and the USA, achieving a standard deviation in the range of about 140 ppm. “Clarity” of Art.2 gets by far the highest variation (sv = 3.5), followed by compatibilities among the three criteria (Fig. 6). The major aim of the survey was not to measure hard “objective” facts, but rather to acquire soft “subjective” data, based on the opinions and perceptions of experts. These can shape political processes and be based on simple assumptions, prejudices and lack of knowledge. Asking for the assumed positions of other actors provides insights about mutual perceptions before actors have identified or stated a clear position publicly. Putting a complex issue on a scale between two extremes facilitates the acquisition and comparison of data but on the other hand may lead to oversimplifications. The more experts know about the complexity and conditionality of the issue, and the more they see everything connected with everything, the more difficult it becomes for them to fix a unique position. In some respect such problems also occur in real-world decisionmaking. Even complex negotiations with many actors and many criteria involved may end with a single agreed number.

20

Figure 6: Standard variation for multi-actor and multi-criteria evaluation

Figure 7: Standard variation for stabilization level and time-frame

21

Future Prospects The negotiations on future international action on climate change will be very complex and cover many diverse and inter-related dimensions. They will address a number of conflicting issues: adaptation vs. mitigation; economic efficiency vs. environmental effectiveness; Kyoto vs. nonKyoto; equity vs. obligations for developing countries. To mitigate these conflicts, a number of proposals have been developed, three of which will be mentioned here (Höhne et al. 2005):  The Triptych approach was originally developed to share emission allowances in the EU for the first commitment period under the Kyoto Protocol. It describes a method to share emission allowances among a group of countries, taking into account different sectors and technologies and national differences in emissions and emission reduction potentials. In each of three major sectors of the economy (energy-intensive industry, power-producing and other domestic sectors) specific criteria are applied to calculate partial allowances for each country, for instance minimum proportions for renewable energy or energy efficiency.  With the Contraction & Convergence (C&C) concept, all countries would agree on a global target level of atmospheric CO2, e.g. 450 ppm, and on a path towards this level (contraction). In a second step, the global emission limit for each year would be shared among all countries, including developing countries, so that per-capita emissions converge by a specific date, e.g. 2050 (convergence). With new scientific findings the defined targets for each country can be reviewed and revised. In a full emissions trading scheme countries in demand of allowances can buy them from other countries that receive excess allowances (e.g. least developed countries). Benefits from resource transfer will be limited to the least developed countries and to the first decades until the target has been achieved.  In the Common but Differentiated Convergence (CDC) proposal per capita emissions of Annex I countries converge to a low level within several decades. Non-Annex I countries converge to this level during the same time period but start when their per capita emissions are a certain percentage above global average and could voluntarily accept “positively binding” targets until then. CDC would ensure stabilization of GHG concentrations, with greater flexibility and acceptability to a wider range of countries. Resolution of the critical issues becomes more urgent as scientific projections and observations demonstrate that the impacts of climate change are becoming real and time for effective action may be running out. Therefore, an acceptable and efficient specification of the ultimate objective has to overcome barriers and conflicts related to different disciplines, interests and responsibilities to allow for bold joint action of the international community. A key contribution from the research community would be to integrate climate, economy and policy models in a modular approach to provide tools for decision-support in negotiations and stakeholder dialogues, in finding agreement, setting targets and taking actions (Fig. 8). One issue is to define future global baseline emission trajectories, based on guardrails of dangerous climate change. The baselines and shares for each region are subject to negotiations to find a balance between optimal and admissible emission paths. The allocation mechanism can include tolerable guardrails for each region and principles of equity and fairness, such as shares equal per capita,

22

proportionate to GDP or emissions. A fully integrated assessment model could provide a tool to help understanding of multi-agent interaction in climate policy and clarify the benefits of cooperation and coalition formation.

Figure 8: Links between climate, economic and policy modules in integrated assessment.

23

Annex: The Questionnaire for the Expert Survey on Art. 2 UNFCCC The following questions were raised in the survey at COP-8, for each of the actors (experts, EU, USA, Russia, G77/China), with the respective variables given: 1. Expected net benefits (BENEFIT): What is the expected net benefit (gains minus losses) of fully achieving the objective of Art. 2 for each of the actors? The score ranges from highly damaging (-5) to highly beneficial (+5). 2. Expected costs for reference stabilization level: Assume a stabilization level of 550 ppm for all greenhouse gases (GHG in CO2 equivalent). a) What is the cost of this goal expected by each actor in average percentage of annual global gross domestic product (GDP-PERCENT)? b) How acceptable would the actors find these costs, ranging from prohibitively expensive (-5) to negligible costs (+5) (COST-ACCEPT)? 3. Clarity of meaning (CLARITY): Do the actors have a clear understanding of the meaning of Art. 2, including the ultimate objective (prevent dangerous anthropogenic interference) as well as the conditions for ecosystems adaptation, food production and economic development? The score ranges from completely unclear (-5) to completely clear (+5)? 4. Consistency of objective: How consistent are the three conditions for the time-frame to achieve non-dangerous stabilization in Art. 2 (ecosystems to adapt naturally; ensure that food production is not threatened; enable sustainable economic development)? For each pair of these conditions (ECONECOL, ECONFOOD, ECOLFOOD) evaluate whether they are highly conflicting (-5) or fully compatible (+5)? 5. Inclusion of equity considerations (EQUITY): Are equity considerations important to the interpretation of Art. 2? The evaluation ranges from completely irrelevant (-5) to highly important (+5). 6. Agenda for Art. 2 implementation: By which year should the full implementation of Art. 2 objectives become a key issue in climate negotiations from the viewpoint of each actor? 7. Preferred stabilization level: Which stabilization level of atmospheric concentration of all greenhouse gases (in ppm CO2 equivalent) do the actors prefer? 8. Required time-frame for stabilization: By which year should the preferred stabilization level given in question 7 be achieved from the viewpoint of each actor?

24

References Allen, M.R. and P.A. Stott, J.F.B. Mitchell, R. Schnur, and T.L. Delworth. 2000. Quantifying the uncertainty in forecasts of anthropogenic climate change. Nature 407, 617-620. Alley, R.B. and J. Marotzke, W.D. Nordhaus, J.T. Overpeck, D.M. Peteet, R.A. Pielke, R.T. Pierrehumbert, P.B. Rhines, T.F. Stocker, L.D. Talley, and J.M. Wallace. 2003. Abrupt climate change. Science, 299 (5615), 2005-2010. Andronova, N.G. and M.E. Schlesinger. 2001. Objective estimation of the probability density function for climate sensitivity. J. Geophys. Res. 106(D19): 22, 605-612. Aubin, J.P. and P. Saint-Pierre. 2004. A Viability Approach to Global Climate Change. In: Haurie, A., Vigier, L. [eds.]. Coupling Climate and Economic Dynamics; Kluwer, Dordrecht. Netherlands. Barnett, J. 2003. Security and Climate Change. Global Environmental Change 13(1): 7-17. Billari, F., and T. Fent, A. Fürnkranz, and J. Scheffran. [eds.] 2006. Agent Based Computational Modelling. Springer/Physica, Heidelberg. Germany. Bruckner, T. and G. Petschel-Held, F.L. Toth, C. Helm, M. Leimbach, and H.-J. Schellnhuber. 1999. Climate change decision-support and the tolerable windows approach. Environmental Modeling and Assessment (4) 217-234. Carbone, J.C., T.F. Rutherford, and C. Helm. 2003. Coalition Formation and International Trade in Greenhouse Gas Emissions. Working paper. University of Colorado, Boulder. http://www2.ncsu.edu/unity/lockers/users/v/vksmith/opportunites/coalition.pdf. Carraro, C., and J.A. Filar. [eds.] (1995). Control and Game-Theoretic Models of the Environment. Annals of the International Society of Dynamic Games. Birkhäuser, Boston. USA. Dessai, S. and M. Hulme. 2003. Does climate policy need probabilities? Tyndall Centre Working Paper, 34. Available from http://www.tyndall.ac.uk/publications/publications.shtml. Edenhofer, O. and K. Lessmann, C. Kemfert, M. Grubb, and J. Köhler (2006) Induced Technological Change: Exploring its Implications for the Economics of Atmospheric Stabilization. The Energy Journal. Special Issue on Endogenous Technological Change and the Economics of Atmospheric Stabilisation, March, 57-108. Eisenack, K., J. Scheffran, and J. Kropp. 2006. Viability analysis of management frameworks for fisheries. Environmental Modeling and Assessment. 11: 69–79. Finus, M. 2001. Game Theory and International Environmental Cooperation; Edward Elgar, Cheltenham. UK.

25

Forest, C.E. and P.H. Stone, A.P. Sokolov, M.R. Allen, and M.D. Webster. 2002. Quantifying Uncertainties in Climate System Properties with the Use of Recent Climate Observations. Science. 295, 113-117. Füssel, M. 2006. Social welfare functions in optimizing climate-economy models: A critical review, submitted to: Journal or Environmental Economics and Management. Giorgi F., and R. Francisco. 2000. Evaluating uncertainties in the prediction of regional climate change. Geophysical Research Letters. 27, 1295-1298. Göbeler, F., and J. Scheffran. Extended Power Values and Dynamic Coalition Formation. pp. 338-340. In: L. Petrosjan, N.A. Zenkevich. [eds.] 2002. Tenth International Symposium on Dynamic Games and Applications. Proceedings Vol. I. International Society of Dynamic Games, St. Petersburg. Russia. Grundig, F., H. Ward, and E.R. Zorick. Modeling Global Climate Negotiations. In: U. Luterbacher, and D.F. Sprinz. [eds.] 2001. International Relations and Global Climate Change. MIT Press, Cambridge. USA. Hare, W. Relationship Between Increases in Global Mean Temperature and Impacts on Ecosystems, Food Production, Water and Socio-Economic Systems. pp. 177-186. In: H. J. Schellnhuber, W. Cramer, N. Nakicenovic, T. Wigley, and G. Yohe. [eds.] 2006. Avoiding Dangerous Climate Change. Cambridge University Press, Cambridge. USA. Haurie, A. and L. Viguier. [eds.] 2005. Coupling Climate and Economic Dynamics. Berlin: Springer, Heidelberg. Germany. Höhne, N., and D. Phylipsen, S. Ullrich, and K. Blok. 2005. Options for the second commitment period of the Kyoto Protocol. Research Report, UBA-FB 000771. ECOFYS, Federal Environmental Agency, Berlin. Germany. Intergovernmental Panel on Climate Change. 1994. IPCC Special Workshop Article 2 of the United Nations Framework Convention on Climate Change. IPCC Working Group II. Washington, D.C. USA. IPCC. 2001. Climate Change 2001: Mitigation. Third Assessment Report. Cambridge University Press, Cambridge. USA. Ipsen, D., R. Rösch, and J. Scheffran. 2001. Cooperation in Global Climate Policy: Potentialities and Limitations. Energy Policy. 29(4): 315-326. Keller, K., B.M. Bolker, and D.F. Bradford. 2004. Uncertain climate thresholds and optimal economic growth. Journal of Environmental Economics and Management. 48 (1), 723-741. Keller, K., G. Yohe, and M. Schlesinger. 2006. Managing the risks of climate thresholds: Uncertainties and information needs. An editorial essay. Climatic Change (in press).

26

Kemfert, C. 2004. Climate coalitions and international trade: assessment of cooperation incentives by issue linkage, Energy Policy, 32 (4), 455-465. Kleinen, T. 2005. Stochastic Information in the Assessment of Climate Change. PhD Thesis. Potsdam University. Germany. Krabs, W., and S. Pickl. 2003. Analysis, Controllability and Optimization of Time-Discrete Systems and Dynamical Games. Lecture Notes in Economics and Mathematical Systems. No. 529. Springer, Heidelberg. Germany. Kriegler, E. 2005. Imprecise Probability Analysis for Integrated Assessment of Climate Change, PhD Thesis. Potsdam University. Germany. Kriegler, E., H. Held, and T. Bruckner. 2006. Climate Protection Strategies under Ambiguity about Catastrophic Consequences. In: J. Kropp and J. Scheffran. [eds.] Decision Making and Risk Management in Sustainability Science. Nova Scientific Publications, NY. USA. Lempert, R.J. 2002. A new decision sciences for complex systems. Proceedings of the National Academy of Sciences. 99, 7309-7313. Mendelsohn, R. 2000. Efficient Adaptation to Climate Change. Climatic Change 45: 583-600. Murphy, J.M. and D.M.H. Sexton, D.N. Barnett, G.S. Jones, M.J. Webb, M. Collins, D.A. Stainforth. 2004. Quantification of modelling uncertainties in a large ensemble of climate change simulations. Nature 430, 768 - 772. Nordhaus, W. D. 1993. Rolling the DICE: An optimal transition path for controlling greenhouse gases. Resource and Energy Economics 15, 27-50. Nordhaus, W. D. and J. Boyer. 2000. Warming the World: Economics Models of Global Warming. MIT Press: Cambridge. USA. O’Neill, B.C, M. Oppenheimer 2002. Dangerous climate impacts and the Kyoto Protocol, Science, 296, 1971-1972. Ott, K. and G. Klepper, S. Lingner, A. Schäfer, J. Scheffran, and D. Sprinz. 2004. Reasoning Goals of Climate Protection - Specification of Art.2 UNFCCC. Federal Environmental Agency, Berlin. Germany. Pershing, J., C. Philibert. 2002. Beyond Kyoto: energy dynamics and climate stabilization. OECD/IEA, Paris. France. Rahmstorf, S. and K. Zickfeld. 2005. Thermohaline circulation changes: a question of risk assessment. Climatic Change. 68 (1-2): 241-247.

27

Rethinaraj, T.S.G. 2005. Modeling Global and Regional Energy Futures. PhD Thesis. ACDIS research report. University of Illinois, Urbana-Champaign. (www.acdis.uiuc.edu/Research/ResReports/RethinarajRR.pdf) Scheffran, J., and S. Pickl. 2000. Control and Game-Theoretic Assessment of Climate Change Options for Joint Implementation. Annals of Operations Research. 97: 203-212. Scheffran, J. Economic Growth, Emission Reduction and the Choice of Energy Technology in a Dynamic-Game Framework. 329-336. In: P. Chamoni, et al. [eds.] 2002a. Operations Research Proceedings 2001. Springer, Heiderlberg. Germany. Scheffran, J. Conflict and Cooperation in Energy and Climate Change. The Framework of a Dynamic Game of Power-Value Interaction. 229-254. In: M. Holler et.al. [eds.], 2002b. Yearbook New Political Economy 20. Mohr Siebeck, Tübingen. Germany. Scheffran, J. Interaction in Climate Games: The Case of Emissions Trading. Pp. 1-18. In: J. Geldermann, M. Treitz [eds.] 2004. Entscheidungstheorie und Praxis in industrieller Produktion und Umweltforschung. Shaker, Aachen. Germany. Scheffran, J., and M. Leimbach. Policy-Business Interaction in Emission Trading between Multiple Regions. pp. 353-368. In: R. Antes, B. Hansjürgens, P. Letmathe [eds.] 2006. Emissions Trading and Business, Springer/Physica, Germany. Scheffran, J. The Formation of Adaptive Coalitions. pp. 163-178. In: A. Haurie, S. Muto, L.A. Petrosjan, and T.E.S. Raghavan [eds.] 2006a. Advances in Dynamic Games. Springer/Birkhäuser, Heidelberg. Germany. Scheffran, J. Tools in Stakeholder Assessment and Interaction. In: S. Stoll-Kleemann, M. Welp [eds.] 2006b. Stakeholder dialogues in natural resources management and integrated assessments: Theory and practice. Springer, Berlin. Germany (in press). Scheffran, J. 2006c. Adaptive Management of Energy Transitions in Long-Term Climate Policy. Computational Management Science (in press). Scheffran, J., and A. Battaglini. 2007. Climate and Conflicts: The Impact of Global Warming on International Security. In: W. Hare, A. Battaglini, W. Cramer, and C. Jaeger [eds.] 2007. Key vulnerable regions and climate change. Oxford University Press, Oxford. USA (sumbitted). Schellnhuber, H.J. and W. Cramer, N. Nakicenovic, T. Wigley, G. Yohe. [eds.] 2006. Avoiding Dangerous Climate Change. Cambridge University Press, Cambridge. USA. Schlesinger, M.E. and J. Yin, G. Yohe, N.G. Andronova, S. Malyshev, and B. Li. Assessing the Risk of a Collapse of the Atlantic Thermohaline Circulation. pp. 37-48. In: H. J. Schellnhuber, W. Cramer, N. Nakicenovic, T. Wigley, and G. Yohe. [eds.] 2006. Avoiding Dangerous Climate Change. Cambridge University Press, Cambridge. USA.

28

Schneider, S.H. 2004. Abrupt non-linear climate change, irreversibility and surprise. Global Environmental Change, Part A. 14 (3), 245-258. Singer, C.E. and T.S.G. Rethinaraj, S. Addy, D. Durham, M. Musik, M. Khanna, B. Kuehl, J. Luo, W. Quimio, K. Rajendran, D. Ramirez, J. Qiang, J. Scheffran, T.N. Tiouririne, J. Zhang. 2006. Probability Distributions for Carbon Emissions and Atmospheric Response. Climatic Change (submitted). Sprinz, D. 2005. Long-Term Policy Problems: Definition, Origin, and Responses. Paper Prepared for the Conference “Illuminating the Shadow of the Future: Scientific Prediction and the Human Condition”. Faculty Research Club, University of Michigan, Ann Arbor, MI. 23-25 Sept. 2005. Stainforth, D.A. and T. Aina, C. Christensen, M. Collins, N. Faull, D. J. Frame, J. A. Kettleborough, S. Knight, A. Martin, J. M. Murphy, C. Piani, D. Sexton, L. A. Smith, R. A. Spicer, A. J. Thorpe, and M. R. Allen. 2005. Uncertainty in predictions of the climate response to rising levels of greenhouse gases. Nature. 433, 403 – 406. Stocker, T.F. 1999. Abrupt climate changes: from the past to the future - a review. International Journal of Earth Sciences. 88, 365-374. Svirezhev, Y.M., W. von Bloh, and H.J. Schellnhuber. 1999. “Emission game”: some applications of the theory of games to the problem of CO2 emission. Environmental Modelling and Assessment. 4, 235-242. Tilmann, D. et al. 2002. Agricultural Sustainability and Intensive Production Practices. Nature. 418. WGBU. 2000. World in Transition: Strategies for Managing Global Environmental Risks. German Advisory Council on Global Change. Springer, Heidelberg. Germany. Weber, M., V. Barth, and K. Hasselmann. 2003. A Multi-Actor Dynamic Integrated Assessment Model (MADIAM) of Induced Technological Change and Sustainable Change Economic Growth. Paper presented at EGS-AGU-EUG Joint Assembly. 6-11 April. Nice, France. Weber, M. 2004. A Multi-Actor Dynamic Integrated Assessment Model (MADIAM). Reports on Earth System Science 3/04. Max Planck Institute for Meteorology, Hamburg. Germany. Wigley, T.M.L., and S.C.B. Raper. 2001. Interpretation of high projections for global mean warming. Science. 293, 451-454. Yohe, G., N. Andronova, and M. Schlesinger. 2004. To hedge or not to hedge against an uncertain future climate. Science, 306, 416-417.

29