On the Post-Mortem Relevance of RCP8.5

UPDATE, 6/9/2026. Zeke Hausfather at ClimateBrink has published a post showing a similar analysis of the change in emission trends before beginning in 2013. he comes to similar conclusions that I do, but his analysis is of course better than mine. If you only have time to read one post, read his.

A lot of misinformation has been circulating in recent weeks concerning the IPCC's decision to remove RCP8.5 from use in the upcoming AR7 report. I agree that the decision is a good one, and I think I can show why below with a simple toy model I created. I think it's fair to say that RCP8.5 is no longer a plausible future for us, given current emission reductions and the shrinking cost of renewable energy. For some very insightful comments from others about the death of RCP8.5, see Zeke Hausfather's post at Climate Brink (also Michael Mann's comments on it here), a factcheck by Carbonbrief, Reto Knutti's comments, and the wonderful information at Climate Action Tracker. But the the usual suspects among contrarian political activists have been portraying the IPCC's decision as a win for "skeptics" and/or as an admission that the IPCC's projections have been fraudulent, such that over a decade of research that used RCP8.5 can now be safely discarded. Nothing could be further from the truth; it's actually an indication that human behavior has followed a different path from the one described in RCP8.5. We have mostly flattened the curve in our growing emissions rates, due to the competitive economics of renewable energy and the implementation of climate policies. However, it appears that there are a large number of bad faith actors, shills and grifters who are either incompetently or deceptively misrepresenting what RCP8.5 was and why it has now been discarded.

The RCPs were designed as representative concentration pathways distinguished by their respective changes in radiative forcing by 2100 caused by increases in greenhouse gas concentrations (partially masked by aerosol pollution). They were not predictions or projections; they are better thought of as conditions (or possible courses of action) under which projections can be made. And these outcomes are affected by two things, only one of which we control:

• Human emissions from fossil fuel combustion and land use change.
• Carbon-cycle feedbacks that affect the airborne fraction of human emissions.

Since 1750, humans have emitted about 750 GtC from fossil fuel combustion and land use change, but of that amount only about 42% (310 GtC) has stayed in the atmosphere. The carbon cycle has taken the rest (440 GtC) in the land and ocean sinks, resulting in greening, ocean acidification, etc. The airborne fraction of our emissions increases with continued emissions, meaning the land and oceans will remove a progressively smaller portion of our emissions. The IPCC's estimate for the change in airborne fraction with continued emissions has actually been conservative; their estimates are on the low end of model experiments. This means that it's possible that the same emission pathway may lead to higher concentrations (and therefore a higher radiative forcing in 2100) than the IPCC expects.

Carbon Brief on Carbon-Cycle Feedbacks

RCP8.5 fits any emission-related course of action that results in 8.5 W/m² radiative forcing by 2100. In 2011, when RCP8.5 was defined, it was a "worst case scenario," representative of the ~90th percentile of no policy concentration outcomes by 2100. In other words, of the possible concentration outcomes of "no climate policy" courses of action, RCP8.5 was designed to represent forcings greater than 90% or more of those outcomes. Following 2011, this scenario has sometimes been wrongly described as "business as usual," and I generally agree with many who have said this is a mischaracterization of RCP8.5. There are many possible concentration outcomes of no climate policy courses of action, and RCP8.5 represented roughly the worst 10% of those outcomes. Furthermore, what is generally perceived as "usual" for us changes as new policies are introduced. What was "usual" in 2011 is no longer "usual" today. However, I think it's possible to show that even in 2011, we were not likely on track to achieve 8.5 W/m² radiative forcing by 2100 (assuming the IPCC's estimates of the resulting airborne fraction of our emissions is correct) and getting to those concentrations would have probably required a ramping up of carbon emissions from the trends found in the previous 30 years. My point here is not to take a stance on whether in 2011 a radiative forcing of 8.5 W/m² was plausibly attainable by 2100; rather, I'm suggesting that we weren't likely on a pace that would attain it.

At the same time, RCP8.5 does represent the political rhetoric of a lot contrarian propaganda, especially the rhetoric coming out of political groups like the CO2 Coalition. William Happer (founder of the CO2 Coalition) and Gregory Wrightstone (its executive director), along with many of their published political tracts, claim that the planet is starved of CO2 and even that we're in a CO2 famine. They claim that we need to increase our carbon emissions to give the planet more "plant food," and the "mild warming" resulting from this (if any at all) would be beneficial to humanity. Wrightstone even published a graphic in his book Inconvenient Facts suggesting that CO2 is dangerously low, and we need to get above 2000 ppm CO2 just to get our CO2 "tank" above a quarter full. As I also show here, here and here, CO2 Coalition dogma is that fossil fuel emissions are saving the planet from a possible CO2 famine and mass extinction. For them, we need more of CO2, not less of it. So RCP8.5 conservatively describes the path of action that the CO2 Coalition wants. I think post-mortem RCP8.5 continues to have relevance in showing the likely results of the future that those espousing this kind of contrarian rhetoric want.

From Wrightstone's Book Inconvenient Facts

To make my case, I set up a toy model based on data from the 2025 Global Carbon Budget that shows emissions increased at rates of 0.14 GtC/yr during the 30 years from 1983-2012 but only by 0.022 GtC/yr from 2012-2024. That change in emission rates is directly attributable to efforts to reduce our carbon emissions. For instance, coal emissions peaked shortly after 2010 but exceeded RCP8.5 up until about 2014 and have been in decline ever since. This is a reversal that the CO2 Coalition does not want.

Even Pielke Jr Shows We Reduced Coal Emissions

I extend each emission trend linearly through 2100 to represent a "business as usual" path (1983-2012) and a "continued climate policy" path (2012-2024). To get a "worst case scenario" (what CO2 Coalition wants) path, I set emissions to follow an emission rate of of 0.18 GtC/yr between 2012 and 2100, or about 13% higher than the 1983-2012 trend. This trend is almost sufficient to reach the "1/8 mark" represented in Wrightstone's graphic, so this is a conservative representation of what the CO2 Coalition claims to be advocating for, whether or not it's actually plausible to achieve. Below I show what results from these three linear projections. At the bottom of this post, I also show a table of the resulting annual and cumulative emissions in 2100 due to each pathway. It's of course an over simplification that the emissions from a growing population and technology use over 88 years following 2012 would result in emissions following strict linear trends, but I'm just using this as a simple model to illustrate a point. 

I used the IPCC's assessment for the airborne fraction of our emissions for RCP7 for my 2012-2024 and 1983-2012 scenarios (56%). I used the airborne fraction from RCP8.5 for the CO2 Coalition scenario (62%). The resulting concentrations in 2100 are also on the table below.

For the radiative forcings associated with these changes in concentrations, I used 1) the IPCC's most recent assessment for the forcings for doubling CO2 at 3.93 W/m² and 2) a preindustrial CO2 concentration of 278 ppm. The 2025 Indicators of Global Climate Change assessed total radiative forcing from human activities in 2024 at 3.10 W/m², and CO2 radiative forcings were about 78% of that total (2.4 W/m²). This may be way oversimplified, but for the sake of this post I'm just assuming that same proportion would continue through 2100, meaning that total radiative forcings would be 1.28x the radiative forcing from CO2 alone. This is what I used to assess the effective RCP value associated with each of these scenarios. The "business as usual" (1983-2012) pathway results in 7.5 W/m². The "continued climate policy" (2012-2024) pathway results in 6.4 W/m². The "worst case scenario" (what CO2 Coalition wants) pathway results in 8.5 W/m² (effectively RCP8.5).

This of course still assumes that the IPCC's assessment of the airborne fraction of our emissions is correct, but we've already seen that their values are conservative. If these are underestimated, then we would see larger increases in radiative forcing by 2100 from the same amount of emissions. My "business as usual" projection from 1983-2012 results in 7.5 W/m² using the IPCC's estimate of an airborne fraction of 56%. If the correct fraction is 75%, then CO2 would reach 900 ppm (within the range of possible carbon-cycle feedback outcomes of RCP6.0) and the same cumulative emissions would result in 8.5 W/m² radiative forcing.  There's an article from CarbonBrief that does a very good job of describing the uncertainties here, and I think it's underappreciated in discussions about whether RCP8.5 was ever plausible. Assuming the IPCC's assessment of the airborne fraction is on target, though, this suggests to me that when RCP scenarios were brand new in 2012, we were on a path that would be best represented as RCP7.5, and we would likely have to ramp up our emissions to achieve about 890 ppm CO2 by 2100 to achieve RCP8.5, though this would be a conservative estimate of the "worst case scenario" that the CO2 Coalition wants. 

To calculate the temperature response associated with each pathway, I used a TCR of 1.9°C. Here's the resulting warming by 2100 for each scenario, with HadCRUT5 plotted for comparison.

The table below summarizes all these results. It shows that my "business as usual" path (1983-2024) would lead to about 3.6°C warming. The desired CO2 Coalition path would lead to about 4.1°C warming, and the "continued climate policy" path (2012-2024) would result in 3.1°C warming. The latter overestimates future warming because there are ongoing efforts to reduce emissions. We should not assume that the 0.02 GtC/yr trend of increasing emissions will continue through 2100. At some point in the near future, we're likely to achieve peak emissions, and the resulting cumulative emissions by 2100 will be much lower. This means that we're likely on a path for less than 3°C warming. 

Projection	Emission	Airborne	2100 CO2	2100 Emissions		2100 Radiative Forcing		2100 Global
Scenario	Trend	Fraction	Amount	Annual	Cumulative	CO2	Total	Warming
2012-2024	0.02 GtC/yr	56%	671 ppm	13.3 GtC	1701 GtC	5.0 W/m²	6.4 W/m²	3.1°C
1983-2012	0.14 GtC/yr	56%	781 ppm	23.1 GtC	2117 GtC	5.9 W/m²	7.5 W/m²	3.6°C
CO2 Coalition	0.18 GtC/yr	62%	890 ppm	27.1 GtC	2317 GtC	6.6 W/m²	8.5 W/m²	4.1°C

The results published by the Climate Action Tracker (CAT) are broadly consistent with my results above. Their scenario resulting from emissions prior to the signing of the Paris Agreement results in about 3.6°C warming in agreement with my simple, toy model for "business as usual." Their "policies and action" scenario results in about 2.6°C warming. This suggests they expect continued implementation of climate policies to result in about 0.5°C less warming by 2100 than is indicated by my linear trend of increasing emissions from 2012-2024.

Update: I found the CAT projections for total annual emissions (in GtCO2) for 2023 through 2100 for their "policies and action" scenario which they project will result in 2.6°C warming by 2100. I converted these values to GtC and assumed an airborne fraction of 56%. This resulted in a total radiative forcing in 2100 of 5.45 W/m² in 2100. Keeping the TCR value I used for my scenarios (1.9°C), I came up with 2.64°C warming by 2100, so I take it that my toy model is using values that are consistent with what is used by CAT.

This suggests that CAT's projections mean that they expect the policy changes and actions between now and 2100 to prevent ~1 W/m² of future emissions from my 2012-2024 projection and save about 0.5°C warming from that projection. Following CAT's scenario will also save ~1°C warming expected by the trajectory set between 1983 and 2012. This of course is progress, but it is not yet good news. We need to keep warming from reaching 2°C above the 1850-1900 mean. And this is what the contrarian rhetoric fails to appreciate. The fact that we're not following RCP8.5 does not mean that we don't have a problem. The problem is just as real with or without RCP8.5, because the path we're on is bad, even if there's another path we're not following that's even worse. And we continue to need drastic reductions in our emissions if we expect to stay below 2°C warming.