To the Editor —

In a recent Commentary, Beck and Mahony1 argue that the development of low-emission scenarios has political implications that go beyond making neutral expert assessments. Beck and Mahony suggest that scenario builders need to anticipate the political impacts of the scenarios, which pose “great challenges to conventional ideas of scientific neutrality”1. We agree with the authors that the way in which scenarios are developed for use in climate science and policy should be revisited, but we believe that scenario builders should rather be more scientific and strive for more neutrality, and that improved methods for scenario development can help to achieve these aims.

Methodological advances made within the futures studies community during the last decade have been largely ignored by the climate change research community, including methods such as cross-impact balance2, scenario diversity analysis3 and scenario discovery4. Cross-impact balance is a method for constructing internally consistent qualitative scenarios by defining and systematically analysing networks of influences among scenario drivers. Scenario diversity analysis is used for generating maximally diverse scenarios in order to best represent the huge number of possible scenarios. In scenario discovery, outputs of scenario-based analysis are used for identifying scenarios relevant to a specific decision-relevant criterion by evaluating consequences of alternatives contingent on future developments. In addition to these three methods, approaches to solution-oriented global environmental assessments are emerging, including formalized methods for stakeholder engagement and deliberation about policy alternatives and their implications5. These new methods and approaches could help scenario builders to be more scientific and neutral in at least three ways.

First, they could help to make scenario-based assessments more systematic by providing well-defined step-by-step procedures for the construction, evaluation and selection of scenarios. The process of scenario development to date — as exemplified by the Special Report on Emissions Scenarios6 and the Shared Socioeconomic Pathways7 — has been largely irreproducible and the scenarios' comprehensiveness is hard to verify. Second, they create greater transparency by allowing people not involved in scenario development to understand and assess both the process and the products, that is, the scenarios. This is important both for scientific reasons as well as to ensure ethical and social legitimacy. The typical approach for developing scenarios employed by the climate change research community follows the so-called intuitive logics model8. This approach, based on 'intuitive group thinking', is not very transparent and has problems of reproducibility. Third, they help to recognize uncertainties and encourage systematic exploration of these, as requested by Beck and Mahony, by providing methods that are designed to analyse a wide range of possible outcomes, including the most extreme scenarios. While the designers of the Representative Concentration Pathways (RCPs) clearly stated that the RCPs should not be interpreted as absolute bounds9, in practice the RCP8.5 scenario has served and still serves that role in most climate projections as well as in impact and adaptation assessments. Moreover, climate projections based on RCP8.5 are commonly misinterpreted as an upper limit10.

As the IPCC increasingly faces pressure to 'open up' for consideration of alternative technological possibilities and policy pathways, as argued by Beck and Mahony, they need to rely on methods that enable more transparent processes of scenario development and more systematic exploration of uncertainties. This will be essential if the IPCC continues to aim to be policy-relevant without being policy-prescriptive.