Please note that the data and information posted under the "Open Process" are for historical information only. They were preliminary information released in 1998-1999 as part of the SRES open process and for use in analysis to be contained in the IPCC Third Assessment Report (TAR). For final data (version 1.1) please go back to home page and follow the link.


IPCC Special Report on Emissions Scenarios (SRES)

Nebojsa Nakicenovic


The Intergovernmental Panel on Climate Change (IPCC) decided at its May 1996 plenary session in Mexico City to develop a new set of reference emissions scenarios as an IPCC Special Report. The Working Group III of the IPCC was charged with the task of appointing the writing team for the Special Report on Emissions Scenarios (SRES). The main objective of the SRES is to review the literature and, based on the outcome, formulate a new set of scenarios to replace the six IPCC IS92 scenarios that are widely used as reference emissions trajectories.

The new set of reference scenarios is intended for use in future IPCC assessments and by the wider scientific and other communities working on impacts of future greenhouse gas emissions and on mitigation measures and policies. Scenarios can provide the emissions profiles as inputs to general circulation models (GCMs) and simplified models of regional climate change. They can provide the reference information required for the assessment of impacts of climate change, such as the level of economic activity in different world regions, rates of technological change, or population growth. The same kind of information, in conjunction with emissions trajectories, can serve as benchmarks for the evaluation of alternative mitigation measures and policies. A new set of reference scenarios, tested for consistency, can also provide a common basis and an integrative element for the IPCC Third Assessment Report (TAR).

In January 1997, the IPCC Working Group III appointed the convening lead author together with the writing team, consisting of 36 members. Six lead authors meetings have been held since that time. The work commenced after the first organizational meeting held at the IPCC headquarters at the World Meteorological Organization (WMO) in Geneva on February 7-8, 1997.

During this meeting the writing team reached consensus concerning the overall work program. It was agreed that the scenario development process would consist of four major components: first, a review of existing global and regional emissions scenarios; second, an analysis of the main scenario characteristics and relationships; third, a formulation of "storylines" as narrative scenario characteristics and development of quantitative "templates" for the new scenarios; and fourth, an "open" scenario modeling process involving feedback from various modeling groups worldwide resulting in the final revisions of the new emissions scenarios. Below is a brief overview of progress on these four work components to date. The writing team’s work plan and expected dates of completion for each work component are attached.

 The first component comprises a unique database that now documents more than 428 global and regional scenarios. This database is accessible at an ftp-site; the password and the address can be provided on request. The objective is to expand the database into a scenario assessment tool in itself. Thus, the writing team would like to include as many scenarios as possible, ranging from global and regional to individual country emissions projections. Information about additional scenarios to be included in the database would be greatly appreciated. This essential part of the literature review is complete, but it will be updated throughout the process.

The second component involves statistical analysis of the scenarios in the database. This process is well underway. The first results were documented in 1997. The authors have analyzed the range of main scenario characteristics (or "driving forces"), such as population, economic development, energy consumption, rates of technological change, GHG emissions, etc. Particular emphasis has been given to analysis of medians and the ranges of scenario distributions for each of the salient scenario characteristics. In addition, the group has started to analyze the relationships among scenario characteristics.

The third component involves the development of the initial set of emissions scenarios. The first step was the formulation of narrative "storylines" to be used as additional information for defining quantitative characteristics of initial scenarios. The work on storylines started at the second SRES meeting at the OECD in Paris on April 13-15, 1997 and has progressed so far that the narrative descriptions are now available as background material for the modeling groups to work on scenario quantifications. In the context of this work, each storyline is basically a short "history" of a possible future development of the key scenario characteristics. The second step involves quantification of developments envisaged in the storylines so that these can be captured in formal models, resulting in initial scenario formulations. In January 1998, four modeling groups that are represented by members of the writing team started to formulate the initial scenarios and test them for consistency in terms of both qualitative and quantitative factors with their models. This exercise will be complete by mid-1998. The result will be a set of quantitative "templates" for new emissions scenarios.  

The fourth component involves the so-called "open" process, which began in June 1998 and will last until December 1998. The template scenarios and general information about SRES activities will be posted on a web-site. A particular effort will be made to invite regional and global modeling groups to submit their own scenarios in the format to be specified by the SRES writing team. In general, three types of submissions would be considered. Scenarios published in the reviewed literature that have not been included in the scenario database, new scenario variants based on common assumptions derived from one of the template scenarios and general suggestions how to improve the work of the writing team posted on the web-site (preferably based on the referenced literature). This will facilitate revision of the template scenarios so that they are consistent with the broader spectrum of modeling approaches and regional perspectives. Once the process is complete, the writing team will consider the comments of outside modeling groups, revise the scenarios as appropriate, and prepare a final report to be submitted for peer review in May 1999. The resulting set of scenarios, after the IPCC review process, would be designated as the new set of IPCC emissions scenarios. The attached work plan shows that an approach with overlapping activities has been chosen to provide inputs for the other groups within IPCC that are commencing work on the TAR.

All of this work has been documented in a series of essays produced by the writing team. Current versions of these essays are called zero-order drafts (ZODs). The ZODs will be peer reviewed and published in some form as background material behind the SRES. They will further appear in a special edition of the journal, Mitigation and Adapatation Strategies for Global Change and will undergo peer review during the summer of 1998. Some of the material presented in these essays will later be used in the SRES report.



In 1990 the IPCC initiated development of its first set of greenhouse gas emissions scenarios designed to serve as the inputs to GCMs and facilitate the assessments of climate change impacts. Two years later, this first set of scenarios was updated by the six so-called IS92 reference emissions scenarios (Leggett et al., 1992 and Pepper et al. 1992). In 1994, the IPCC evaluated the usefulness of the IS92 scenarios; this review was published in 1995 under the title, An Evaluation of the IPCC IS92 Emission Scenarios (Alcamo et al., 1995).

The IPCC evaluation summarized the following purposes of emissions scenarios: first, to provide input for evaluating climatic and environmental consequences without specific mitigation measures and policies to reduce emission of greenhouse gases; second, to do the same but with specific policy interventions directed at reducing future emissions, such as mitigation measures and strategies; third, to serve as input in the assessment of mitigation potentials and costs in different regions and economic sectors, and over time; and fourth, to provide input for negotiating possible emissions reductions (Alcamo et al., 1995).

In January 1997, the IPCC Working Group III appointed the convening lead author together with the writing team of 36 members for the Special Report on Emissions Scenarios (SRES). The SRES writing team is fortunate to include members who participated as convening and lead authors in all three of earlier IPCC scenario activities: the first set of scenarios developed in 1990, the IS92 scenarios developed in 1992 and the 1994 scenario evaluation. The SRES activities are intended to build upon earlier IPCC emissions scenario work and thus assure continuity in the overall assessment. The main tasks of the SRES team are to review prior scenarios from the literature and to develop a new set of IPCC reference emissions scenarios.

The 1994 IPCC evaluation of the IS92 scenarios found that the scenarios were innovative at the time of their publication and that they were path-breaking in that they covered the full range of GHGs emissions as well as sulfur dioxide both on a global and a regional basis. For example, they provide aggregate sulfate aerosol emissions for the world as a whole as well as the regional emissions profiles. The review also identified a number of weaknesses of the IS92 scenarios, including a narrow range of carbon intensities of energy (carbon emissions per unit energy) across all six scenarios. Furthermore, the IS92 scenarios obviously could not account for the changes that have occurred since their inception, such as the decline in CO2 emissions resulting from economic reform and restructuring in the former Soviet Union and Eastern Europe. Moreover, the IS92 scenarios used estimates for the base year 1990 as the actual data were not available at that time. Today, 1990 is no longer a forecast year, since actual base year data are available up to the year 1995.

The advent of integrated assessment (IA) models in the meantime has made it possible to construct internally-consistent emissions scenarios that include interactions between emissions of GHGs and other pollutants, such as sulfate aerosols, emissions mitigation, and possible impacts of climate change. Progress has also been made in achieving greater consistency among various scenario characteristics and features, such as the rate of technological change across different sectors.

Owing to these advancements, it has become necessary to develop a new, updated set of reference emissions scenarios. This brief paper describes key elements of the new scenario development process and documents its progress to date.


To ensure maximum utility for future work of IPCC and other scientific assessments, new reference scenarios should meet several criteria. They should cover the full range of emissions, including GHGs and other radiatively active pollutants, such as sulfates and particulates. This is particularly important because the extent that the new scenarios will be used as input for making climate change projections within GCMs and other regional climate impacts models. The scenarios should also have sufficient spatial resolution to allow regional disaggregation. In particular, high priority should be given to consistency with the proposed regional assessments of the new Working Groups of the IPCC (Watson 1996). New scenarios should be rich in the sense of spanning a wide spectrum of alternative futures, structural changes and salient uncertainties.

Meeting these criteria will require a set rather than one single scenario. However, to be useful, the scenarios should not be too numerous. Rather, a minimum set should be identified that is sufficiently rich in range. Currently, the writing team is considering a range of four scenarios with a larger set of variants. What is more important than the number per se is that scenarios differ in their underlying structure. Differences between scenarios should not be achieved merely by varying a single characteristic, such as the underlying population projection or energy consumption. Instead, they should reflect a range of different underlying structures, as reflected in quantitative scenario characteristics and qualitative "storyline" descriptions. In turn, those different structures will define a range of different assumptions for the key "driving forces", such as population, economic development, technological change, etc. Some scenarios should be neutral with respect to policies and measures that may be taken to slow global warming. This is consistent with the approach that was taken in developing IS92 scenarios.

All scenarios will need to include assumptions about other measures and policies that are not pursued in response to global warming, but which could affect climate change. For example, one scenario could envision drastic reductions of sulfate aerosol emissions in response to concerns about "acid rain"; another could strive to close the North-South gap; others might include the objective of eradicating poverty or achieving freer trade in the world.

Ideally, the writing team would propose a classification scheme for scenarios that utilizes structural features as an aid in determining the minimal scenario set and the individual scenario characteristics. To develop a useful scheme it may be necessary to look beyond existing methods. For example, attempts to attach probabilities to different scenarios have not led to any major insights, so that any such simplistic criterion for scenario classification should be avoided. The number of possible scenarios is practically infinite, so that any classification scheme is to an extent arbitrary. Whatever the final choice will be it is important the new set of scenarios covers the range of major underlying "driving forces" of emissions scenarios identified in the SRES literature review.

The scenarios need to be suitable for integrated assessment by including the "whole loop" that ranges from human activities to GHG emissions, atmospheric concentrations, climate change and impacts, and the most important feedbacks within the loop. They need to be tested as to whether they are suitable for a wide range of possible uses, especially those relevant for future IPCC assessments.

Perhaps most crucial are the requirements that the new IPCC scenarios be transparent, reproducible, and internally consistent. These criteria can be fulfilled by ensuring that assumptions and results are documented and also stored in a database and an appropriate web-site. The combination of extensive literature review, development of coherent scenario "storylines" and formulation of each scenario by a number of different modeling approaches can help to ensure internal consistency. Moreover, the scenario formulation process should be open to the scientific community and other groups involved in such activities. This will also enhance pluralism of methods and approaches from different disciplines and parts of the world. Conventions for reporting results and documentation should be standardized to ensure comparability across scenarios.


The authors of the SRES met for the first time at IPCC headquarters at the World Meteorological Organization (WMO) in Geneva, Switzerland on February 7-8, 1997 to fully define their objectives. Specifically, they identified the requisite characteristics of the new scenarios and created modeling teams for scenario formulation. They also agreed to establish contacts with relevant users, including various ongoing IPCC assessments, such as the Special Report on Aviation and the Global Atmosphere and the Special Report on Regional Impacts of Climate Change.

The second authors meeting took place at the United Nations University in Tokyo, Japan on March 15-16, 1997. At this meeting, participants discussed the literature relevant to the development of new scenarios. These topics included population, resources and energy supply, end use technologies, land use changes, structural shifts, climate feedbacks, sulfur emissions, methane, and other greenhouse gases. The team agreed to review the literature on these issues comprehensively in a series of essays that will be peer reviewed and published separately as a special issue of a scientific journal. They also discussed the design of a database containing scenarios from the literature and reviewed progress on collecting such scenarios. Analysis of the range of emissions found in outside sources will enable the team to develop a set of scenarios which adequately reflects the current thinking in the scientific and other communities.

The authors met for the third time at the OECD in Paris, France on April 13-15, 1997. Here the emphasis was on reaching a common understanding of possible storylines for the new IPCC emissions scenarios. Storylines offer qualitative descriptions of the assumptions and relationships underlying each of the scenarios, thereby providing a credible rationale for the scenarios’ respective emissions trajectories. Thus, storylines constitute an important element of the scenario development process.

The fourth authors meeting took place at IIASA in Laxenburg, Austria on June 14-16, 1997. Participants reviewed progress on the zero-order drafts and further refined the storylines based on additional insights gained from the ongoing literature review. They further began the quantification of key relationships and variable distributions for the scenarios. Also on the agenda were presentations by IPCC experts who employ emissions scenarios in climate modeling and in regional impacts assessments. Insights from the user community will contribute to the development of scenarios that correspond to the needs of future users.

On September 17-19, 1997 the writing team met again at RIVM in Bilthoven, the Netherlands, to present the latest versions of the review essays. Completion of these reviews is a prerequisite for the final formulation of scenario storylines, initial scenario quantification and consistency assessment of the scenarios through a limited set of formal models, and commencement of the open process mandated for the SRES. The authors also discussed progress on storylines, analyses of scenario drivers, and preliminary scenario quantifications. Progress to date has been substantial and is in line with the tentative timetable. Particularly advanced are the literature surveys and the compilation and analysis of scenario database. The group further revised the work plan for the coming months and fleshed out the details of the open process, which will begin next year.

An informal authors meeting was held February 7-8, 1998 at Lawrence Berkeley National Laboratory in Berkeley, California, USA. The main purpose of this meeting was to review the progress of the modeling groups that have been involved in first quantifications of the four storylines. Participants discussed technical issues, storyline interpretation and consistency of first quantifications, emphasizing harmonization of results. It was agreed that the first round of model runs would be produced and circulated by March-April, prior to the next lead authors meeting. The group also reported on the progress of the zero-order drafts as well as the possibility of publishing the ZODs in a journal. It has since been determined that the ZODS will appear in a special issue of Mitigation and Adaptation Strategies for Climate Change later this year. Finally, the writing team discussed the open process, which began in June 1998 and will end in December 1998. CIESIN (The Consortium for International Earth Science Information Network) in the United States will manage the open process, creating an IPCC-SRES web-site to collect the input of interested outside parties.

The entire writing team met again at the World Bank in Washington, D.C. on April 29-30, 1998 to assess the progress of the modeling and to further discuss the ZODs. The modeling groups discussed standardization of scenario quantifications and agreed on "marker" scenarios for each scenario family. The marker scenario is the one scenario that the writing team believes provides the best quantitative interpretation of the qualitative scenario description. It represents the modelers’ best attempt to give users a single consistent set of numbers. While only one case will be the marker, all other quantifications will be retained to show variations of scenario quantifications.

In addition to harmonizing model inputs, the writing team also reviewed the ZODs and discussed arrangements for publication in the special issue of Mitigation and Adaptation Strategies for Climate Change. A representative of CIESIN addressed the group regarding the details of the open process, including web-site contents and design, publicity, submission requirements, and reporting of final results. On the second day of the meeting, Bob Watson, chairman of the IPCC, presented an overview of the upcoming IPCC Third Assessment Report (TAR). This was followed by a discussion of the Working Group III's contribution to the TAR and how the new scenarios could contribute to the TAR.


The writing team has agreed that the scenario development process will consist of four major components: first, a review of literature on global and regional emissions scenarios; second, an analysis of the characteristics and relationships among major driving forces in the scenarios from the literature surveys; third, a formulation of "storylines" as narratives of scenario characteristics and development and testing of "templates" for new scenarios by a few modeling groups; and fourth, an "open" scenario modeling process involving feedback from various regional modeling groups. Below is a brief overview of progress on these four work components to date. Attached is the updated work plan of the writing team that gives expected dates of completion for each work component and other related activities.

Review of Literature and Existing Scenarios

A thorough literature review is a prerequisite for evaluating the state-of-the-art and documenting progress in emissions scenarios achieved since the evaluation of the IPCC IS92 scenarios (Alcamo et al. 1995). It is also a prerequisite for the open process envisaged for the SRES and for the documentation of the ranges across different regional and global emissions scenarios and their underlying assumptions. The literature review and documentation of global and regional emissions scenarios will extend previous assessments and build on the 1994 IPCC Scenario Evaluation (Alcamo et al. 1995) to include recent research findings (e.g., IIASA-WEC study; SEI-UNEP scenarios; Shell scenarios) and contributions from other groups involved in such assessments, such as the Energy Modeling Forum (EMF), International Energy Workshop (IEW), and Energy Technology Systems Analysis Programme (ETSAP).

The review activity consists of two elements. One is the review of literature and existing scenarios on all of the key driving forces that are relevant for the development of scenarios such as population, economic development, technological change, availability of energy resources, etc. These efforts, which will culminate in the publication of peer reviewed essays on each of the key scenario driving forces, will update and revise the review conducted for the last IPCC scenario evaluation (Alcamo et al. 1995).

The other element of this review activity is a comprehensive review of assumptions and results from existing scenarios. One of the primary instruments to be used as part of this process is a database compiled by Dr. Tsuneyuki Morita and his colleagues at the National Institute for Environmental Studies in Japan. This database contains over 428 scenarios submitted to the writing team to date or collected from literature and other efforts, such as EMF and IEW. By performing statistical analyses of the scenarios in the database, the IPCC authors can ensure that the new scenarios fully reflect the range of input assumptions and emissions projections currently found in the open literature. In the coming year more scenarios will be added to the database to ensure that it truly reflects the state-of-the-art in GHG emissions modeling.

Analysis of Main Scenario Characteristics and Key Relationships

The systematic review of the existing literature and scenarios will make possible analysis of the key scenario characteristics and relationships. Such analysis is required for establishing benchmarks that can be used as a starting point for the development of new scenarios that adequately reflect the range of values in the current literature and scientific knowledge. This process involves assessment of important relationships – such as between economic growth, demographics, technological change, and energy consumption. Key characteristics and relationships for the new scenarios will be determined by the SRES writing team, depending on the outcome of the ongoing literature review. Basing the key characteristics and relationships among the scenario driving forces on the current literature, together with the use of models to formulate scenarios, is a way to ensure that the new scenarios are internally consistent.

The ranges and variations of main scenario driving forces and characteristics are being captured in the form of histograms, distribution functions, and associated statistics, such as the mean and the percentiles. The sample means from the literature review and analysis can be used as one set of typical benchmarks. For some variables and characteristics the samples from the literature and scenarios reviews may be too small to derive meaningful statistics solely from the existing literature.

The required analysis has been divided into topical groups, each to be coordinated by a member of the SRES writing team. For example, subgroups will address such issues as population, resources and technology, end-use technologies, land-use changes, structural change and shifts, climate feedbacks, non-CO2 GHGs, SOx , CH4, and other GHGs, such as N2O, and CFCs.

Development and Testing of Scenario Narratives and Quantifications

Each scenario will be backed by a credible narrative description in the form of a storyline, or a qualitative description of the scenario’s "future" history. The purpose of the storylines is to describe main scenario characteristics and relationships among key driving forces. Another purpose of these storylines is to provide descriptive background information to aid in the quantification of scenario assumptions. It is a description of a scenario that goes beyond the numbers. Yet another function of storylines is to help with the interpretation of scenarios, e.g., to give causal reasons for improvement rates of energy intensities in relation to rates of technological change, economic growth, and GHGs emissions across different scenarios.

The process of storyline formulation must be creative and imaginative, but it also requires consistency checks between (alternative) narrative storylines with quantitative scenario assumptions, driving forces, and benchmarks derived from the literature review and assessment. This will require both a subjective element, based on alternative scenario storylines, and a quantitative aspect based on the literature and scenarios reviews.

Once the storylines have been completed, they will need to be quantified into template scenarios. An important step in this process will be the selection of the appropriate modeling frameworks to test the initial template scenarios. Several integrated assessment modeling groups will then be engaged to test the initial templates numerically in their models and assist in revision of the main assumptions, characteristics and results until consistency is achieved. Discussions have been initiated with modeling groups at the Pacific Northwest National Laboratory (PNNL) in the U.S.; the National Institute for Public Health and the Environment (RIVM) in the Netherlands; the National Institute for Environmental Studies in Japan (NIES); the International Institute for Applied Systems Analysis (IIASA) in Austria, and the IEA’s Energy Technology Systems Analysis Programme (ETSAP) network. Standardized scenario reporting forms and data exchange protocols will be developed by the writing team in order to facilitate communication between modeling groups. All of these groups are represented on the SRES writing team.

The result of this testing process will be a set of qualitative storylines and quantitative templates that are internally consistent. Together they will constitute the initial set of new scenarios. In may ways, the storylines and templates could provide the required input to the work of other IPCC groups, such as the impacts community. They would have many features of the final scenario set, such as the range of demographic and economic development, or the emissions of greenhouse gases and sulfate aerosols.

Open Scenario Modeling Process

The development of the new scenarios will be conducted by the writing team through an open process involving modeling groups from around the world and in coordination with other IPCC writing teams working on modeling climate change, impacts, and mitigation. CIESIN has offered to host a web-site for this purpose. The web-site will include information on IPCC, SRES, the nature of the open process, and SRES template scenarios. It will also invite submission of alternate scenarios. There will also be links to and from other web-sites as well as a help facility to answer user questions. All interested researchers and organizations will be invited to evaluate with their modeling frameworks the template scenarios and the relevant assumptions. Groups will be encouraged to make such evaluations for the world as a whole, for some world regions, or individual countries of their choice. They will be asked to modify the template scenarios as they see appropriate, using standardized scenario reporting forms and data exchange protocols. Upon completion of the open process in December 1998, the writing team will then analyze the submissions and integrate the relevant findings into the set of new scenarios and recommend this final set for adoption as the new IPCC emissions scenarios.

The open process should involve groups from different parts of the world and groups that represent different methodological approaches. It would be desirable if the open process included modeling groups that employ different methodological approaches, including systems-engineering and macroeconomic models, as well as a number of so-called integrated assessment models. The pluralism of methods would test whether the template scenarios can fulfill the likely needs of different user communities within the IPCC and groups that have contributed to the study.

This open process will help in achieving two important objectives for developing new emissions scenarios. First, it will allow inclusion of the most recent findings in the revised (and final set of) scenarios the continuation of scenario comparison and evaluation with potential user groups. Second, it will help encourage the active involvement of individual researchers and experts, various research organizations, as well as relevant stakeholder groups, such as government institutions, intergovernmental organizations, industry, environment groups and other non-governmental organizations (NGOs).

Timing of SRES Activities

The IPCC Third Assessment Report (TAR) is expected to be completed by mid-2001 (Watson, 1996). The timing of Special Report on Emissions Scenarios is critical, given the schedules for TAR and other on-going activities in the IPCC work program. Clearly, much of the work for SRES and TAR will need to proceed in parallel, owing to the limited time available. For example, new climate change projections with GCMs are needed as input to the TAR impacts assessments. These are computationally very demanding, take a long time to complete, and require emissions scenarios as input. Thus, the whole process cannot be conducted sequentially and still be complete in 2001.

Nevertheless, the September 1996 IPCC Working Group I workshop in London concluded that it may be possible to use the new emissions scenarios to run simple climate models within this short time period. The new emissions scenarios could be used that way directly for the impacts assessment without necessarily involving the more elaborate GCM runs. The same probably applies for the assessment of mitigation and adaptation measures and strategies.

To be useful in this process, the SRES would need to be completed, reviewed and approved by IPCC in 1999. The authors will follow a process that involves parallel and overlapping activities, which will involve developing the new set of scenarios and at the same time coordinating the open process to be followed by the IPCC evaluation and review process. A more detailed SRES schedule indicating the timing of major work components and activities can be found in the appendix, as can a list of contributing authors and the terms of reference.



Alcamo, J.; A. Bouwman; J. Edmonds; A. Gruebler; T. Morita; and A. Sugandhy. 1995. "An Evaluation of the IPCC IS92 Emissions Scenarios," in Climate Change 1994: Radiative Forcing of Climate Change and an Evaluation of the IPCC IS92 Emission Scenarios: 247-304. Cambridge University Press.

Leggett, J.; W. Pepper; R.J. Swart; J. Edmonds; L.G. Meira Filho; I. Mintzer; M.X. Wang; and J. Watson. 1992. "Emissions Scenaros for the IPCC: An Update." In Climate Change 1992: The Supplementary Report to the IPCC Scientific Assessment, University Press, Cambridge, UK.

Pepper, W.; J. Leggett; R. Swart et al. 1992. Emission Scenarios for the IPCC an Update: Assumptions, Methodology, and Results. Cambridge University Press. Cambridge, UK.

Watson, R. 1996. "The IPCC Third Assessment Report and the IPCC Bureau." Discussion Paper.