Thoughts about scenarios for mitigating climate change – a commentary on the Scenarios Forum 2019 (for the IPCC’s 6th Assessment Report)

A guest commentary by Dr. Richard A. Rosen (Tellus Institute, Cambridge, MA – retired) 

The Scenarios Forum 2019 is a conference that will be held at the University of Denver from March 11-13, 2019.  This forum aims to bring together a diverse set of research communities working within the various frameworks regarding integrated climate change mitigation scenarios to share their experiences, progress and plans. The frameworks use Shared Socioeconomic Pathways (SSPs), Representative Concentration Pathways (RCPs) and other approaches to characterizing future societal and environmental conditions to investigate global climate change issues, including the related SDGs. By taking stock of progress and facilitating scenario-related research, this meeting will inform the use of scenarios in the IPCC Sixth Assessment Report (AR6) and help ensure a research base sufficient to inform future climate change mitigation assessments and policy initiatives.  The thoughts expressed below describe why a much broader view of the concept of a scenario needs to inform any future work input to the IPCC’s Working Group III report for AR6 for those scenario to be policy relevant.


A scenario is a reasonably internally self-consistent set of quantitative and qualitative assumptions that combine to make a story line into a potentially quantifiable projection for the future.  Of course, for the purpose of quantifying or modeling many features of a scenario to mitigate climate change, the systemic complexity of any scenario is so great as to make it very difficult to judge if the assumptions which define any given scenario are internally consistent or not.  Much more empirical research is typically required to make these judgements than is typically used as a basis for creating mitigation scenarios.  This typical research deficit makes it very difficult to know which scenarios are more or less plausible than others that have been published in the past in the peer-reviewed literature.  This problem of what is a reasonably self-consistent set of assumptions which could comprise an interesting scenario for mitigating climate is partially due to the fact that most if not all of the Integrated Assessment Models that the IPCC has primarily relied on in the past for creating scenarios are not at all sufficiently documented so that the internal consistency or not of the scenarios run on these models can be determined.  For example, it is rarely clear how the major investments required to mitigate climate change will change the economies of different regions.  And it is never discussed how the financing of mitigation investments will work, among other issues not discussed in research articles that report findings from IAMs.


As implied above, a scenario is not just a few sentences or paragraphs describing a possible future, but a well thought out scenario includes all of the assumptions that must be specified which allows that scenario to be run in an IAM or similar model.  For example, if a scenario is meant to be the basis for a future as computed by an integrated assessment model, as many of the scenarios relied on by the IPCC in its assessments typically are, then such a scenario should explicitly include every numerical input assumption, and every equation, used to compute the results of such a scenario by these models.  In fact, numerical input assumptions such as the capital costs of new nuclear plants are even more important than the scenario story line that goes with a scenario since they more directly determine the characteristics of that scenario in the future.  Note that many published results for the same SSP run on different IAMs are radically different from each other in part because the capital cost assumptions of new energy technologies are significantly different from each other.  Yet, those differences are rarely discussed in the literature.  (Scenario inputs made public also must include items like the numerical value of discount rates and other key assumptions that may be appropriate across an entire set of scenarios.) Thus, as has unfortunately been the case in past IPCC assessments, if the equations, constraints, and numerical values of all input assumptions are not transparent to the scientific community, and to the public, who are trying to understand the significance or implications of such scenarios, then achieving the necessary level of understanding for climate change mitigation policy making is impossible.  For example, it is never pointed out to policy makers that the single most important assumption for policy making in each scenario is the discount rate used in IAMs which minimize discounted costs and benefits.  This is because the use of different discount rates will produce radically different results in most if not all IAMs.  Since Working Group III has recently committed itself to greater research transparency than has been the case with the published research literature in the past, this transparency must be understood as essential to the proper explication and presentation of all mitigation scenarios relied on in AR6.  The revised SSPs, which should be an output of this Scenarios Forum, should be completely specified and completely transparent because the mitigation policy community will be able to put any faith in their usefulness for AR6


All AR6 scenarios that are finally relied on for publication in the literature, and in the final AR6 reports, should have detailed and comprehensive input from a multi-stakeholder community, and not just from the IAM modeling teams themselves, as has been the situation in the past.  Thus, this Scenarios Forum should be the beginning of that process of creating viable AR6 scenarios, not the end of that process. The IPCC can not assume that all or even most of the policy relevant and interesting mitigation scenarios will just by chance appear in the published literature based on the separate and private inclinations of the IAM modeling community to run those scenarios through their models.  The IPCC working group meetings, especially the WGIII meetings in addition to this Scenarios Forum called by IAM modelers to discuss scenarios, must have substantial input from other stakeholders, such as environmental and social justice organizations, into the issue of which scenarios are ultimately modeled, and, perhaps even which models are used to run them, to the extent that different models may have different capabilities to model certain kinds of scenarios.


No, the SSPs are not “true” or complete scenarios, because when they are run through different IAMs and are translated by different research teams into different numerical input assumptions, they produce very different results for the future for what is currently called the same SSP.  Thus, it is very misleading to readers when different research papers written by groups of authors using different IAMs report dramatically different results that they call “for the same scenario”, just because their model runs are loosely based on the same SSP.  In the future, if this is done, readers need to be clearly told that the results of different IAMs for the same SSP do not represent the same scenario at all for the future of climate change mitigation.  Furthermore, when the SSPs are described in “research” papers, it is not even clear whether or not the same numerical values for the key input parameters described in words in those papers are required to be used for each different IAM.  (ref to original SSP papers)  I believe that in fact the same numbers have not been required to be used for the same input parameters by different IAM research teams in the past.  Thus, we all need to be clear that if one of the purposes of using the SSPs for the WGIII AR6 report is to be able to correctly claim that different models are being used to calculate the results for the same SSP, then two things need to happen.  First, revised SSPs and fully comprehensive need to specify the precise numerical value to be assigned to each of the hundreds of input parameters that each model uses, and the different sets of internal model constraints and equations that have a significant impact on the production of any given SSP-based scenario must be made transparent and public so that the differences between model results for the so-called “same” SSP can be understood and explained, to the extent possible, by each modeling group.  In the past, SSPs have only described a few of the key input assumptions that differ between scenarios.  This problem must be corrected in the future, hopefully beginning at this Scenarios Forum.


In past IPCC WGIII climate change mitigation assessments, it has not been at all clear what value has been provided to policy makers by referencing the published literature involving the running of different Integrated Assessment Models for creating mitigation scenarios.  Frankly, policy makers might learn more if many research teams had collaborated to build one “consensus” model where all the features of the model were made public, including all equations and input assumptions.  The way research in the IAM community has evolved, most if not all the IAMs that appear frequently in the literature are accompanied with little or no technical documentation of their equations and numerical input assumptions.  Nor can adequate documentation of the models be found on the research institution websites.  To policy makers the models all appear to be “black boxes” with no way to understand the basis for the results of scenarios that emerge and are published in research papers.  This extremely serious problem for policy makers must be corrected, in my view, in AR6.  Otherwise, there is no point and no value gained from doing another IPCCWGIII report if the literature on which it is based is not improved in the ways described above.

In addition, the next generation of IAMs used for research inputs to AR6 must be able to model different climate change mitigation policies beyond just modeling carbon taxes, otherwise how will policy makers be able to understand the possible real world impacts of other kinds of policies that will be essential to successfully mitigate climate change?  Examples of other policies that need to be included in at least some revised SSPs are regulatory/legal requirements to phase in electric vehicles at different rates in different regions of the world, and regulatory requirements (RPSs) to phase in renewable electricity supplies at different rates in different regions of the world.  Other policies that also need to be able to be modeled include the phase in of organic food, and the regulated (required) phase in of electric space heating, cooking, and hot water in both residential and commercial buildings throughout the world.  Similarly, different rates of energy efficiency gains for building shells should be included in different SSPs.

If these kinds of essential mitigation policies are able to be modeled in different groups and combinations, then policy makers may finally get some IAM model results that might truly help them make mitigation policy decisions as soon as possible after AR6 is published.  Of course, this all should have happened in the past in the literature published in the process of preparing for AR5 by modeling teams.  As all policy makers should know by now, the four main components of all plans for mitigating climate change to either a 1.5 or a 2.0 degree C future are:  make all energy end use technologies as efficient as reasonable given their social context, electrify all energy end-uses to the extent possible, produce all electricity using renewable electricity supply technologies, and produce all food supplies from organic and sustainable agricultural processes.  If all aspects of these four basic and high priority mitigation policies cannot be modeled by the IAMs utilized to create the research literature for AR6, then the WGIII report for AR6 will not be able to heavily rely on IAM results as it has in the past.  To be very clear, then, different revised SSPs created for use in AR6 must include different combinations and magnitudes of the different key mitigation policies listed above in order to produce different alternative relevant scenarios for use by policy makers.


This scenarios forum is very important in order for the world climate change mitigation policy community to have some confidence that the WGIII AR6 report will, indeed, be much more policy relevant than previous IPCC WGIII reports have been.  Other than attempting to model the impact on GHG emissions using carbon taxes (unsuccessfully in my view), almost no attention has been paid to modeling the impact of other kinds of legal and regulatory mitigation policies that have been most successful in the past.  The time available to the world to sufficiently mitigate climate change is rapidly growing short.  The mitigation research community cannot afford to delay any more in bringing the policy community more precise, scientifically-based, and useable mitigation policy analyses.  We now know after the IPCC Special Report on 1.5 degree C Scenarios from October, 2018 that almost all, if not all, GHG emissions have to be mitigated by about 2045 at the latest, if no overshoot of 1.5 degrees C is desired.