Introducing the Representative Emission Pathways

A new community proposal for emissions pathways for the upcoming IPCC AR7 cycle

Emissions scenarios provide an important underpinning of the work that climate scientists do to assess future warming outcomes and impacts. They represent one of the first of many steps needed to produce climate model outputs that inform the IPCC assessment process.

But emissions scenarios also represent a moving target; to be useful to policymakers and the broader public we need to explore scenarios that span a range of plausible outcomes – including worlds where no additional action is taken to mitigate climate change or where emissions mitigations are rapidly undertaken to try and meet our most ambitious climate goals. As such, emissions scenarios are a frequent source of controversy, with researchers debating the plausibility of different assumptions both on the high and low end of the range.

In a new preprint available for community review in Geoscientific Model Development (GMD) a diverse set of researchers have proposed developing a new set of Representative Emissions Pathways (REPs) to underlie the climate modeling work in the upcoming IPCC AR7 cycle. These REPs are intended to be used by climate models to create new Representative Concentration Pathways (RCPs) and Representative Warming Pathways (RWPs), which can be used to explore Representative Hazard Pathways (RHPs), Impact and Risk Pathways (RIPs), Adaption Pathways (RAPs), Socioeconomic Pathways (e.g. SSPs), and Transition Pathways (RTPs) as illustrated in the figure below.

Conceptual overview of climate change scenarios, representative emission pathways (REPs) and the framework climate pathways. Figure 1 from Meinshausen et al 2023.

Separating the world that is from the world avoided

Emissions pathways utilized by the IPCC process need to serve a number of purposes. They should explore both high-end emissions scenarios where no additional action is taken on climate change, while also assessing a wide range of mitigation outcomes ranging from low to high ambition. At the same time, these should largely reflect the world as it is today, rather than assuming a hypothetical time machine where we could take more ambitious action or undo existing climate progress that has already been achieved.

To that end, we propose that new REPs tie more directly to the literature on mitigation outcomes associated with both current policies and commitments made by countries than has been the case in past IPCC cycles. We separately encourage researchers to undertake a lower priority effort to assess worlds avoided: both higher-end warming outcomes that seemed more plausible a decade ago, and worlds where emissions peaked in 2015 to enable us to confidently limit warming to 1.5C with no overshoot. However, we stress that these latter pathways should be explicitly seen as worlds avoided – counterfactuals with which to compare currently plausible emissions scenarios – rather than realistic outcomes given where we are today.

The four recommended high priority REPs include:

• NFA: “No Further Action”, a category for a pathway reflecting current emission futures in the absence of any further climate action, with warming of around 2.5-3.0C by 2100. The pathways in this category should ideally be accompanied by a perturbed physics ensemble as this would allow us to obtain valuable proxies for a worst-case high-end warming outcome under emissions implied by current policies (e.g. low likelihood / high impact outcomes) and avoid the need for researchers to use less plausible high-end emissions scenarios as an imperfect proxy.

• DASMT: ‘Delayed climate Action and Stabilisation pathway Missing Target’, a category for a pathway that misses the Paris Agreement long-term temperature goal as it results in global warming of around 2C in 2100, rather than staying “well-below” 2C. Such a pathway explores global emissions approximately in line with NDCs and longterm targets as they were proposed around the time of Glasgow, COP26, (approximately resulting in a median just below 2C warming, if fully implemented) .

• DAPD: ‘Delayed Action Peak and Decline’, a category for a pathway in which climate action is further delayed, but that then features rapid emission declines and strongly negative long-term CO2 emissions.

• IAPD: ‘Immediate Action Peak and Decline’, a category for a pathway that features immediate 2025 onset of decisive emission reductions and achieves net zero CO2 emissions by mid-century.

In addition, we propose two lower-priority “worlds avoided” and “opportunities missed” pathways:

• TEWA: ‘The Emission World Avoided’, a category for a pathway with high or very high emissions analogous to SSP3-7.0. This would allow a depiction of the world that could have unfolded without climate policies and help provide a high signal-to-noise ratio for projected changes in climate, though researchers should ensure that it is clearly communicated that this reflects a world avoided rather than a plausible emissions pathway today.

• IA2015: ‘Immediate Action in 2015’, a category for a pathway that resembles 'a world that could have been at the low emissions end, assuming emission reductions towards net-zero had started in 2015. Other world that could have been scenarios can be envisioned, and could be policy relevant, e.g., one starting in 1992 with the establishment of 565 the UNFCCC, or 2009.

The figure below shows an illustrative example of the CO2 emissions and global mean temperature increase across each of these pathways:

Overview of framing pathway categories for Earth system model runs. Figure 2 from Meinshausen et al 2023.

Recognize that the world does not end in 2100

For decades the climate science community has used 2100 as a cutoff date for our climate model runs. This is somewhat motivated by the computational cost of running earth system models on high performance supercomputers, but there is nothing particularly significant about the year 2100 apart from it being a nice round number.

As we’ve gotten closer to the end of the century – with people alive today (including my daughter) likely living well into the 22nd century – its becoming increasingly clear that ending climate model runs in 2100 is increasingly untenable. As such, we’ve proposed extending the modeling period for both emissions scenarios and climate model runs through the year 2150, with some select pathways (e.g. NFA) extended through 2500 for modeling groups that are able to do longer timescale runs.

Better reflect uncertainties in carbon cycle feedbacks

When we explore future climate outcomes we have to deal with three different uncertainties: our future emissions, the sensitivity of the climate to our emissions, and the response of the carbon cycle to our emissions. The latter one reflects how carbon cycle feedbacks change as the planet warms, and how those changes affect the portion of our CO2 emissions that remains in the atmosphere.

One of the limitations of the last set of climate model runs done in the IPCC AR6 ScenarioMIP process (e.g. the main future scenarios featured in the report) is that they effectively exclude carbon cycle feedback uncertainties from future warming projections. Because a small number of modeling groups still lack biogeochemical cycles in their climate models (and rather focus on modeling atmospheric physics and ocean dynamics), they are unable to translate emissions of CO2 and other greenhouse gasses to radiative forcing and warming.

To include as many modeling groups as possible, the IPCC AR6 continued the approach that prior IPCC assessment reports had used to run climate models using prescribed concentrations of different greenhouse gases and other forcings, rather than running models based on emissions. These prescribed concentrations were based on a best-estimate of carbon cycle feedbacks, but lacked any uncertainty across different modeling groups. A separate CMIP experiment – C4MIP – explored uncertainties across models using emissions driven runs, but it only included a small number of modeling groups and a single scenario (SSP5-8.5).

For the AR7 cycle it is time for a change. We suggest that ScenarioMIP runs should all be emissions driven – at least for CO2 – providing a more robust estimate of uncertainty in future warming across different emissions pathways.

An invitation for community commentary

This preprint is the start of a broader community discussion around the emissions pathways to be used in the IPCC AR7 cycle. Its by no means a final decision, and we welcome and encourage researcher to leave comments as part of the GMD open review process. Deciding what emissions pathways to use will be a critical decision that shapes much of the work done by climate scientists over the next decade, and the more input we have in this process from different parts of the scientific community the more robust the resulting pathways will be.

Related posts

RCP8.5 asks: why can't we just get along?