"Neither Desirable, nor Possible"

Part 3: Groupthink, misinformation and a failure of self-correction

Over the past week I have documented in two previous posts how the Intergovernmental Panel on Climate Change (IPCC) made a consequential error in its Fifth Assessment Report (AR5) in 2013 in identifying RCP8.5 as the only legitimate baseline scenario for climate research and assessment.

Today in a third and final installment, I provide my explanation for the error — groupthink fueled by a misinformation campaign led by activist climate scientists. What I cannot explain is why the error has not since been corrected by the IPCC or others in authoritative positions in the scientific community. In fact, the opposite has occurred — RCP8.5 remains commonly used as a baseline in research and policy, perhaps explaining why it continues to have many defenders.

The graph below shows the minimum radiative forcing¹ in the scenarios of the IPCC designated as baselines — where we thought we might be headed.² The graph shows how baseline scenarios became much more extreme over a very short period during the drafting of the IPCC’s Fifth Assessment Report.

In 2000 the IPCC used 40 baseline emissions scenarios to project futures with a wide range of possible climate outcomes, with the lowest having a radiative forcing of about 4.5 Watts per meter-squared (W/m²) and the highest >8 W/m². By 2014, with the publication of the AR5 the IPCC had focused on a single baseline scenario — RCP8.5.

In 2012 and 2013, the IPCC AR5 collected scenarios that had been developed following the publication of the 2007 AR4 report.³ Most of these 1,184 scenarios were produced by a small group of integrated assessment modelers who participated in the Energy Modeling Forum Model Inter-comparison Project 27 (EMF 27). The four RCP scenarios were drawn from those scenarios developed using these various models.

Across these many models was a shared assumption of global “renaissance in coal” through the 21st century. You can see the implications of that shared assumption in the figure below from the work of Justin Ritchie who has meticulously and quantitatively documented the flawed assumptions underlying the IPCC AR5.⁴

In the upper right of the figure above you can see the extreme RCP8.5 and SSP5 scenarios that the IPCC identified as baselines. As you can see, these scenarios assume a 7-8x increase in coal consumption by 2100, implying the building of >30,000 new coal power plants this century — a rate of more than one per day coming online every day until 2100. Implausible.

The figure above also provides a clear indication of groupthink in the shared assumption that coal consumption would increase dramatically — an assumption grounded in a theory that has proven incorrect. The consequences of the erroneous projection of a “coal renaissance” was magnified by the identification of the extreme RCP8.5 as the only legitimate baseline within the broad range of expectations for increasing coal consumption.

It is important to understand the global context in which these assumptions were made. In the early 2000s carbon dioxide emissions were increasing rapidly, driven by development in China powered largely by coal. A non-expert might have thought that a trend observed over several years might accurately represent energy systems over the next 90 years, but scenario experts knew better.

Even so, misinformation took root. A widely cited paper in PNAS⁵ claimed that the carbon dioxide emissions growth rate from 2000 to 2004:

was greater than for the most fossil-fuel intensive of the Intergovernmental Panel on Climate Change [SRES] emissions scenarios developed in the late 1990s

That claim was false. It was made by climate scientists, not energy systems experts.

The false claim that emissions across all SRES scenarios had already been exceeded was widely repeated and promoted, making its way into a 2009 United Nations climate science report. Some climate scientists advanced the idea that that the IPCC scenarios were already out of date and not extreme enough.

By 2010, scenario experts became so frustrated about the misrepresentation of their work that 18 of them, many leaders in the field, wrote a letter to Nature titled, Misrepresentation of the IPCC CO2 emission scenarios, calling out climate scientists for misrepresenting the IPCC, explaining:

contrary to some statements in recent publications current emissions are not higher than covered in the [SRES] climate change scenarios used by the last two Intergovernmental Panel on Climate Change (IPCC) assessments

However, the letter did not change the momentum of misinformation and groupthink.

What happened next was the ratcheting up of expected emissions in baseline scenarios as the IPCC AR5 report was being drafted.

Here is a timeline that goes along with the bar chart found above in this post.

• 2011: The RCPs are published, RCP4.5 is identified as a baseline (i.e., 4.5 W/m²)⁶

• 2012: The IPCC AR5 first order draft is written. It says: “virtually all baseline scenarios 4 lead to radiative forcing above 5.0 W/m2 by the end of the century”

• 2013: The IPCC AR5 final draft is written. It says that the majority of baselines are “above 6.0”

• 2014: The IPCC AR5 is published and RCP8.5 is the only RCP scenario within the baseline range.

In 2023 the IPCC warned that RCP8.5 should not be used as a reference scenario and a group of experts convened to start preparing a next generation of scenarios have largely left RCP8.5 behind.⁷

Even so, RCP8.5 continues to have champions. For instance, in a response to a paper that I wrote with Justin Ritchie, IPCC AR5 Working Group 2 co-chair Chris Field⁸ and Marcia McNutt, President of the U.S. National Academy of Science wrote:

. . . the high-emissions RCP8.5 scenario has long been described as a “business-as-usual” pathway with a continued emphasis on energy from fossil fuels with no climate policies in place. This remains 100% accurate . . .

One important characteristic of science is that it is self-correcting — over time we expect to get closer and closer to reliable knowledge. Mistakes and misunderstandings are correctable, and it is our collective commitment to correction that advances understandings.

On climate scenarios, self-correction has failed.

In 2000, the IPCC presented 40 baseline scenarios that described an envelope of possible emissions futures that in 2023 we still remain within — in fact, near the center. By 2014 we had discarded those scenarios and decided to focus research and policy on a single extreme baseline — now known to be misleading, flawed and not fit for purpose.

Back in 2000 the SRES authors warned us:

The broad consensus among the SRES writing team is that the current literature analysis suggests the future is inherently unpredictable and so views will differ as to which of the storylines and representative scenarios could be more or less likely. Therefore, the development of a single “best guess” or “business-as-usual” scenario is neither desirable nor possible.

Today, RCP8.5 is deeply woven into the fabric of climate research and policy. Getting back on track will not be easy or without opposition. Understanding how we got here should provide a cautionary warning for how science can go astray when we allow self-correction to fail.

I have appreciated your comments and pointers as I have taken this deep dive into climate scenarios. Please like and share all three parts of this exploration of how it is that the IPCC got so off track. Thanks for subscribing!

1. The higher the level of radiative forcing the greater the projected climate change.

2. Acronyms and source links: SRES = Special Report on Emissions Scenarios, AR4 = IPCC Fourth Assessment Report, AR5 = IPCC Fifth Assessment Report, FOD = First Order Draft, SOD = Second Order Draft

3. The discarding of AR4 and earlier scenarios was clearly a bad decision. It is not clear why this was done.

4. Ritchie’s work, in collaboration with Hadi Dowlatabadi is among the most significant research on climate of the past 20 years. Everyone should read his classic paper, Why do climate change scenarios return to coal?.

5. The PNAS paper was cited ~1,400 times from 2007 to 2014.

6. One curiosity is that the lead author of the paper that introduced RCP4.5 as a baseline was also a Lead Author of the IPCC Chapter that excluded RCP4.5 as a baseline.

7. Extreme scenarios do have valuable roles to play in exploratory research.

8. Of note, Field was a co-author of the PNAS article cited above that introduced the false claim about 2000-2004 emissions rates exceeding the range of the SRES scenarios.