A part of the problem is us scientists. We use jargon no one understands, and we equivocate over insignificant details. The willful deniers and the head-in-the-sand deniers use our own language to convince the public that there is no problem.
"The second point of concern, involving resilience, covers several potential surprises in the Earth system response to such warming: current Earth system models may lack or underestimate the strength of some important climate-related positive feedbacks under these high-temperature conditions. The warming response could be stronger than commonly anticipated. These cases are conceptually sketched as purple arrows in Figure 2. They are an attempt to address the open question to what extent the Earth system will conserve its Holocene resilience under human pressures. Put another way, to what degree will the Earth system continue to physically and biologically buffer Anthropocene stress (i.e., to which degree it continues to be constantly dominated by negative feedbacks to warming that either take energy out of the system (like the Planck radiation feedback) or greenhouse gases out of the atmosphere (like CO2 fertilization)?
Thank you — that's a section of the paper I didn't use, and it's a fair addition.
If sensitivity is state-dependent, a comparison at fixed three and four degrees understates the warm cases. On a pathway that stays hot, sensitivity climbs; on one that comes back down, it doesn't. So the gap between the mitigated and unmitigated pathways would be wider than the fixed comparison I show, not narrower — which runs in the same direction as the argument rather than against it. Treating it as fixed is the conservative choice, and I should have said so. And your way of putting it is the right one: a sensitivity that rises with temperature is a feedback in its own right, not a parameter.
I made a version of this point in the loop essay using Tierney's work, where sensitivity is already elevated at Pliocene concentrations. Anagnostou is the Eocene version, from boron isotopes. I appreciate the engagement and emphasis.
Hi Dean, well said! For me as a non-expert, your compilation is particularly useful. I wonder whether the name "Equilibrium Climate Sensitivity" is adding to the confusion. What is meant by equilibrium is not well defined and maybe an artificial construct that is not actually occurring in nature. An earlier post from you laid this out very nicely. It is really ESS or even AESS that deal with the real Earth climate equilibrium. So, using ESS or AESS might be easier to communicate although it does put things into pretty far future
You mentien that your projected path is still far from your equilibrium band. Again, the fact that the Earth's climate doesn't move towards the equilibrium band is difficult for me to accept. It would imply that the time constant to equilibrate is much longer than the time constant for carbon removal. But this would mean that CO2 behaves similar to methane, but on a longer time scale, and Earth could never have warmed up except during short periods of about 100,000 years after volcanic eruptions. This is obviously not true. Either your carbon removal time is too short or the move towards equilibrium is much faster. My vote would be for 500 year time with ESS = ~ 7 K and longer with AESS = ~ 8 - 9 K (your band).
Thomas — you're right that the system returns to equilibrium. I think we only disagree about the route.
You're expecting the trajectory to climb toward the band. What the figure shows is the other way round: carbon dioxide comes down, and the system rejoins the band near where it left it, at something close to preindustrial concentrations. Same destination, approached from the other side.
And the beginning of that route isn't mine. The turn and the first part of the descent are in the published runs — CLIMBER-X, and the ScenarioMIP projections themselves. All of them turn over and start back down within a few decades of net zero. What my model adds is only how long the rest of it takes.
On the wider point, I'd put it differently than you have. Over the last 66 million years the Earth has mostly been cooling — from Eocene warmth above a thousand parts per million down to the ice ages. The warm intervals were held up by continuing volcanic supply, not by one pulse lingering. That is the difference from methane: carbon dioxide can be resupplied on tectonic timescales. Take the supply away and it cools.
My issue is not with your path extension. It is with the model to 2500. I think, over these 500 years it should have gotten to ESS sensitivities and be much closer to the equilibrium. Only with this faster relaxation to equilibrium can I make sense of the close tracking of temperature and CO2 during the ice ages.
BTW the lowering of CO2 at the end of the model path to 2500 is due to CO2 absorption in the oceans and reaching an equilibrium between oceans and atmosphere. This is not CO2 removal. After 2500 the CO2 concentration should stay fixed as shown in Clark et al.
My comment about your path was just to argue that if it were true that the temperature/CO2 is back to the starting point after just 10 - 30 ka the Earth would never warm up in the first place since volcanic eruptions were typically spread out over hundreds of thousands of years. Now thinking about this, I would then say that your carbon removal is probably too fast. I understand that weathering takes millions of years.
You may be right about the rate, and I should say so plainly. Clark and colleagues report sixty to seventy per cent of peak warming remaining after ten thousand years; my model gives thirty-three. A factor of two too fast, and I say so in the companion piece.
But it doesn't hold fixed after 2500. Between ocean uptake saturating and silicate weathering there's a third stage — carbonate compensation, seafloor calcium carbonate dissolving to neutralise the excess — and it runs on thousands of years, not millions. So the concentration keeps falling, more slowly than before and faster than weathering alone would manage.
Which may mean we're closer than it looks. Too fast in my model, but not fixed either.
I'm a geologist so I look to the past when trying to predict the future. It seems like time for a deep dive into ultra-high-resolution analysis of the Paleocene-Eocene Thermal Maximum. Glad I still have my library access!
Dean, that's the best synopsis of leading climatologists' ECS derivations!
Whether 3,4,or 5C it is catastrophic even now because the carbon sinks are quickly diminishing as net zero gets renamed to lame duck drawdown
The CO2 will double faster rather than slower.
It it looks looks, walks, and quacks like a duck it sure is one.
Even the lower ECS temps will be reached faster ( even if they are mistakingly low ) and the urgency is not the final ECS number but the quickness we reach it by destroying Earth's carbon sinks.
And, I feel the water cycle disruption is very under valued/rated in many assessments of speed of temp rise in ECS derivations.
What would the “case for acting” look like, given that if we did simply burn all remaining fossil fuel reserves, anthropogenic CO2 emissions could only rise to ~700ppm, well below a doubling from today’s 430ppm to 860ppm as assumed in the models discussed in this essay?
Are 8.3 billion humans going to suddenly stop burning high energy density fossil fuels and voluntarily collapse modern industrial civilisations back to pre industrial population numbers of say 1 or 2 billion?
Remember we can’t build and maintain low energy density harvesting machines (wind and solar) using electricity plus we’re running out of many critical input materials like copper, helium, fertilisers, uranium, plastics, rare earth metals etc. all needed to maintain 8.3billion of us.
Take your number as given. Burn everything and you reach about 700 parts per million. That's my central case — the Medium pathway here peaks at 693 — so we already agree about where this ends.
Where we part is what follows. At that concentration the planet sits around four degrees above preindustrial and stays within a couple of tenths of it for thousands of years, because carbon leaves the air far more slowly than it goes in. That follows from the amount of carbon, not from who burned it or why.
So consider what your scenario hands forward. Future generations get the heat, for millennia. And they get it with the fossil carbon spent, because that is what burning all of it means. They inherit the consequences and none of the options.
That's the asymmetry. We have the reserves and the decision. They would have neither.
And your materials argument cuts the other way. If copper and rare earths are genuinely scarce, that's a reason to spend them building something durable while we still can — not a reason to burn the carbon first and leave the next century with a hot planet and nothing to build from.
Two things worth separating. The durability of specific machines and the exact technology mix — that's engineering detail, and reasonable people are still working it out.
But the direction isn't in doubt. We have to move to lower-carbon energy eventually, whatever the timeline. If civilization is going to persist, that answer has to be found — burning the remaining carbon doesn't remove the requirement, it removes the resources for meeting it. And the energy system studies I've read suggest that with the right mix, all of the above rather than any single technology, we get a long way toward net zero. Not easily, and not with one answer. But a long way.
Which is the point about timing. Doing it now means doing it with fossil energy still available to build with. Doing it later means doing it on a hot planet with whatever's left. Same transition either way — only the conditions change.
Engineers and supply chain analysts and geologists tell us that on this Pale Blue Dot there’s nowhere near enough fossil energy (nor raw materials which can’t be mined or transported or manufactured in many critical process with electricity) to build solar energy flow harvesting machines, at anywhere near the scale needed to replace the ~80% of global energy that fossil fuels have been providing humans (which has remained the same since the 1970s in other words electricity remains at only 20% of global energy and wind & solar are still only 3.1%)
"If a reader’s sense of urgency rises and falls with where that estimate lands, then something has gone wrong in how the case was made — and those of us making it share the responsibility."
Dean, THAT statement is the EPITOMY of the problem. We are on 🔥 NOW and ECS SHMECS knowledge is not putting out the fire or recognizing the fire now.
A part of the problem is us scientists. We use jargon no one understands, and we equivocate over insignificant details. The willful deniers and the head-in-the-sand deniers use our own language to convince the public that there is no problem.
We Are in the Anthropocene—Now What? Johan Rockström, , Stefan Rahmstorf
First published: 06 June 2026 https://doi.org/10.1029/2025EF007730
"The second point of concern, involving resilience, covers several potential surprises in the Earth system response to such warming: current Earth system models may lack or underestimate the strength of some important climate-related positive feedbacks under these high-temperature conditions. The warming response could be stronger than commonly anticipated. These cases are conceptually sketched as purple arrows in Figure 2. They are an attempt to address the open question to what extent the Earth system will conserve its Holocene resilience under human pressures. Put another way, to what degree will the Earth system continue to physically and biologically buffer Anthropocene stress (i.e., to which degree it continues to be constantly dominated by negative feedbacks to warming that either take energy out of the system (like the Planck radiation feedback) or greenhouse gases out of the atmosphere (like CO2 fertilization)?
"
#Rockström 46 8
A state dependent ECS (with higher sensitivity at higher temperatures) can therefore be
regarded as a positive feedback acting independently of current human forcing. .
surprises in the Anthropocene.
https://www.nature.com/articles/s41467-020-17887-x
https://publications.pik-potsdam.de/pubman/faces/ViewItemOverviewPage.jsp?itemId=item_34521
https://bsky.app/profile/thomas-boettcher.bsky.social/post/3mnwyw7yics2z
Thank you — that's a section of the paper I didn't use, and it's a fair addition.
If sensitivity is state-dependent, a comparison at fixed three and four degrees understates the warm cases. On a pathway that stays hot, sensitivity climbs; on one that comes back down, it doesn't. So the gap between the mitigated and unmitigated pathways would be wider than the fixed comparison I show, not narrower — which runs in the same direction as the argument rather than against it. Treating it as fixed is the conservative choice, and I should have said so. And your way of putting it is the right one: a sensitivity that rises with temperature is a feedback in its own right, not a parameter.
I made a version of this point in the loop essay using Tierney's work, where sensitivity is already elevated at Pliocene concentrations. Anagnostou is the Eocene version, from boron isotopes. I appreciate the engagement and emphasis.
Hi Dean, well said! For me as a non-expert, your compilation is particularly useful. I wonder whether the name "Equilibrium Climate Sensitivity" is adding to the confusion. What is meant by equilibrium is not well defined and maybe an artificial construct that is not actually occurring in nature. An earlier post from you laid this out very nicely. It is really ESS or even AESS that deal with the real Earth climate equilibrium. So, using ESS or AESS might be easier to communicate although it does put things into pretty far future
You mentien that your projected path is still far from your equilibrium band. Again, the fact that the Earth's climate doesn't move towards the equilibrium band is difficult for me to accept. It would imply that the time constant to equilibrate is much longer than the time constant for carbon removal. But this would mean that CO2 behaves similar to methane, but on a longer time scale, and Earth could never have warmed up except during short periods of about 100,000 years after volcanic eruptions. This is obviously not true. Either your carbon removal time is too short or the move towards equilibrium is much faster. My vote would be for 500 year time with ESS = ~ 7 K and longer with AESS = ~ 8 - 9 K (your band).
Again, your graph is very thought provoking!
Thomas — you're right that the system returns to equilibrium. I think we only disagree about the route.
You're expecting the trajectory to climb toward the band. What the figure shows is the other way round: carbon dioxide comes down, and the system rejoins the band near where it left it, at something close to preindustrial concentrations. Same destination, approached from the other side.
And the beginning of that route isn't mine. The turn and the first part of the descent are in the published runs — CLIMBER-X, and the ScenarioMIP projections themselves. All of them turn over and start back down within a few decades of net zero. What my model adds is only how long the rest of it takes.
On the wider point, I'd put it differently than you have. Over the last 66 million years the Earth has mostly been cooling — from Eocene warmth above a thousand parts per million down to the ice ages. The warm intervals were held up by continuing volcanic supply, not by one pulse lingering. That is the difference from methane: carbon dioxide can be resupplied on tectonic timescales. Take the supply away and it cools.
My issue is not with your path extension. It is with the model to 2500. I think, over these 500 years it should have gotten to ESS sensitivities and be much closer to the equilibrium. Only with this faster relaxation to equilibrium can I make sense of the close tracking of temperature and CO2 during the ice ages.
BTW the lowering of CO2 at the end of the model path to 2500 is due to CO2 absorption in the oceans and reaching an equilibrium between oceans and atmosphere. This is not CO2 removal. After 2500 the CO2 concentration should stay fixed as shown in Clark et al.
My comment about your path was just to argue that if it were true that the temperature/CO2 is back to the starting point after just 10 - 30 ka the Earth would never warm up in the first place since volcanic eruptions were typically spread out over hundreds of thousands of years. Now thinking about this, I would then say that your carbon removal is probably too fast. I understand that weathering takes millions of years.
You may be right about the rate, and I should say so plainly. Clark and colleagues report sixty to seventy per cent of peak warming remaining after ten thousand years; my model gives thirty-three. A factor of two too fast, and I say so in the companion piece.
But it doesn't hold fixed after 2500. Between ocean uptake saturating and silicate weathering there's a third stage — carbonate compensation, seafloor calcium carbonate dissolving to neutralise the excess — and it runs on thousands of years, not millions. So the concentration keeps falling, more slowly than before and faster than weathering alone would manage.
Which may mean we're closer than it looks. Too fast in my model, but not fixed either.
Excellent post. The point about it being highly relevant to remaining carbon budgets is well made.
I'm a geologist so I look to the past when trying to predict the future. It seems like time for a deep dive into ultra-high-resolution analysis of the Paleocene-Eocene Thermal Maximum. Glad I still have my library access!
"We are on the eve of a great crisis." John Quincy Adams
Dean, that's the best synopsis of leading climatologists' ECS derivations!
Whether 3,4,or 5C it is catastrophic even now because the carbon sinks are quickly diminishing as net zero gets renamed to lame duck drawdown
The CO2 will double faster rather than slower.
It it looks looks, walks, and quacks like a duck it sure is one.
Even the lower ECS temps will be reached faster ( even if they are mistakingly low ) and the urgency is not the final ECS number but the quickness we reach it by destroying Earth's carbon sinks.
And, I feel the water cycle disruption is very under valued/rated in many assessments of speed of temp rise in ECS derivations.
What would the “case for acting” look like, given that if we did simply burn all remaining fossil fuel reserves, anthropogenic CO2 emissions could only rise to ~700ppm, well below a doubling from today’s 430ppm to 860ppm as assumed in the models discussed in this essay?
Are 8.3 billion humans going to suddenly stop burning high energy density fossil fuels and voluntarily collapse modern industrial civilisations back to pre industrial population numbers of say 1 or 2 billion?
Remember we can’t build and maintain low energy density harvesting machines (wind and solar) using electricity plus we’re running out of many critical input materials like copper, helium, fertilisers, uranium, plastics, rare earth metals etc. all needed to maintain 8.3billion of us.
Take your number as given. Burn everything and you reach about 700 parts per million. That's my central case — the Medium pathway here peaks at 693 — so we already agree about where this ends.
Where we part is what follows. At that concentration the planet sits around four degrees above preindustrial and stays within a couple of tenths of it for thousands of years, because carbon leaves the air far more slowly than it goes in. That follows from the amount of carbon, not from who burned it or why.
So consider what your scenario hands forward. Future generations get the heat, for millennia. And they get it with the fossil carbon spent, because that is what burning all of it means. They inherit the consequences and none of the options.
That's the asymmetry. We have the reserves and the decision. They would have neither.
And your materials argument cuts the other way. If copper and rare earths are genuinely scarce, that's a reason to spend them building something durable while we still can — not a reason to burn the carbon first and leave the next century with a hot planet and nothing to build from.
Thanks Dean. What machines can we build that’s durable beyond a few decades? And where will we get the energy to build and power them?
Two things worth separating. The durability of specific machines and the exact technology mix — that's engineering detail, and reasonable people are still working it out.
But the direction isn't in doubt. We have to move to lower-carbon energy eventually, whatever the timeline. If civilization is going to persist, that answer has to be found — burning the remaining carbon doesn't remove the requirement, it removes the resources for meeting it. And the energy system studies I've read suggest that with the right mix, all of the above rather than any single technology, we get a long way toward net zero. Not easily, and not with one answer. But a long way.
Which is the point about timing. Doing it now means doing it with fossil energy still available to build with. Doing it later means doing it on a hot planet with whatever's left. Same transition either way — only the conditions change.
Engineers and supply chain analysts and geologists tell us that on this Pale Blue Dot there’s nowhere near enough fossil energy (nor raw materials which can’t be mined or transported or manufactured in many critical process with electricity) to build solar energy flow harvesting machines, at anywhere near the scale needed to replace the ~80% of global energy that fossil fuels have been providing humans (which has remained the same since the 1970s in other words electricity remains at only 20% of global energy and wind & solar are still only 3.1%)
enegyskeptic.com
thehonestsorcer.substack.com
simonmichaux.com
Media and Scientist should talk about chances
ECS= 3° there's a 27% chance
ECS= 6° there's a 10% chance
ECS 6K is wipe out of civilization.
fat tail risk
( stop talk probability ) @djspratt.bsky.social 5
https://www.youtube.com/watch?v=n6ocu-R9IOs&t=740s
https://bsky.app/profile/thomas-boettcher.bsky.social/post/3mqbsafohfs2j
https://bsky.app/profile/thomas-boettcher.bsky.social/post/3msdehuki2c2m
"If a reader’s sense of urgency rises and falls with where that estimate lands, then something has gone wrong in how the case was made — and those of us making it share the responsibility."
Dean, THAT statement is the EPITOMY of the problem. We are on 🔥 NOW and ECS SHMECS knowledge is not putting out the fire or recognizing the fire now.
As a practicing alclimist I always assumed ECS=pi C.
Being convinced of Hansen's value I adjusted mine to the same factoring in water.
+1 C = 7% more humidity so
Pi C + 1.7C = 4.8 C My magnum dopus 🤣