Thanks for sharing this paper. I disagree with most of the key claims you make here.
Firstly, you cite the recent updates on climate sensitivity, but without clarifying that this has narrowed uncertainty at both the left and right of the tail. “The most recent estimates of equilibrium climate sensitivity show a similar distribution, narrowing the range of outcomes to exclude rises below 2 °C but not ruling out warming above 4.5 °C (Sherwood et al 2020).” The first thing to say is that equilibrium climate sensitivity measures the warming conditional on an increasing in co2 concentrations—it doesn’t tell us the unconditional probability of a particular level of warming. Moreover, this is misleading as a description of the Sherwood et al (2020) paper because it makes it sound like the right tail hasn’t changed. The paper also excludes ECS above 5.5 degrees.
Secondly, you say “GHG concentrations of 700 ppm could produce a 10% chance of exceeding a temperature rise of 6 °C (Wagner and Weitzman 2015). Such concentrations would be passed by 2100 under six of the nine of the CMIP6 SSP-RF baseline and 6.0 W/m2 forcing scenarios (Riahi et al 2017, Gidden et al 2019).” As Johannes and I discuss in our recent post, this risk now seems a lot lower
Thirdly, this is not a criticism, but the Raftery estimates have since been updated by Liu and Raftery who find “The median forecast for 2100 is 2.8 °C, with likely range (90% prediction interval) [2.1, 3.9] °C. The median is 0.4 °C lower than that of Raftery et al.1, the upper bound is 1.0 °C lower, while the lower bound is 0.1 °C higher. The tighter interval reflects the additional 5 years of data and the improved model.”
Fourthly, I’m not sure whether I have misunderstood your paper, but the entire impacts literature focuses on the impact of >3 degrees of warming. The reason for this is that it focuses on the impacts of RCP8.5 at 2100, which implies warming of 4-5 degrees. It seems like in your search strategy, you have searched for mentions of particular levels of warming. But if you had searched for mentions of RCP8.5, you would have found that almost the entire impacts literature focuses on comparing that scenario to RCP2.6 or lower.
It is true that our phrasing around Sherwood et al (2020) can sound a bit misleading. However, this was not in bad faith. We did not intentionally leave out the decreased tail risks.
Overall, your criticism mainly seems to be the fact that current estimates of climate sensitivity have a smaller range that the one we used. This is true. It does not change the point of our paper though. While the range decreased, the mean basically stayed the same and if you look at our figures, this still means that there is likely to few research for the higher temperatures.
To your last point, I wasn’t aware that the impact literature mainly compares those two scenarios. However, as we look at all parts of the IPCC reports this seems to get drowned out by the other parts of the reports. We are currently doing some new analyses that try to look into this in a bit more depth.
On the Sherwood et al thing, another issue is that I don’t understand why you would use the Weitzman estimate of ECS when you have already mentioned the Sherwood et al (2020) estimate. The Sherwood et al (2020) estimate is superior, and that is clear from reading their paper. The IPCC also now accepts that it is superior. So, I don’t understand why you would mention “the most recent estimate” and then use one from 6 years ago instead.
“Overall, your criticism mainly seems to be the fact that current estimates of climate sensitivity have a smaller range that the one we used.” That is not my criticism. My criticism is that emissions look set to lower and the right tail of climate sensitivity has thinned which makes a clear material difference to your conclusions.
On your last point: this does make it look like the central claim of your paper is wrong. I am going to paste some charts from the last impacts report below on some key climate impact areas to illustrate:
Sea level
Food yields
Floods
Species loss
Fire frequency
Ocean net primary production
Precipitation
Maximum fish catch potential (A1B is between RCP6 and RCP8.5)
On drought they say “Climate change is likely to increase the frequency of meteorological droughts (less rainfall) and agricultural droughts (less soil moisture) in presently dry regions by the end of the 21st century under the RCP8.5 scenario (medium confidence). {WGI AR5 Chapter 12} This is likely to increase the frequency of short hydrological droughts (less surface water and groundwater) in these regions (medium evidence, medium agreement).”
The Sherwood reference was only included during the review process, as it was not yet published when we originally came up with the analysis. As you probably know, going from an idea to a published paper can take quite some time and you cannot read and update on all the papers that are published during that time.
I would agree that today the picture looks better as in comparison when we started working on that paper. However, predicted temperatures and mentions in the IPCC still don’t overlap and therefore we still have a research gap, albeit a smaller one.
I don’t see how this makes the paper wrong. 1) You are only referring to impact literature. It’s great if we know what impact climate change will have, but we still have a problem if we do not know how to mitigate it at higher temperatures. 2) Comparing emission scenarios is good, but every scenario has a wide range of possible temperatures. Therefore, it is also important to look at specific temperatures.
I’m not sure I understand point (1) in your last paragraph. ‘Mitigation’ with respect to climate usually refers to reducing CO2 emissions, so I don’t see how there could be a specific problem of ‘mitigating at higher temperatures’. Perhaps you mean adapting to higher temperatures?
I don’t think your point (2) hits home. The impacts literature that I have outlined and pasted above does look at specific temperatures. It looks at the most likely level of warming on RCP8.5.
In your paper you say that warming of >3 degrees is ‘severely neglected’. This is not true, and hasn’t been for years. If you had said >5 degrees that would have been true, but the claim is about >3 degrees.
That’s what I meant, sorry if I phrased this incorrectly.
I did not mean to say that they did not look at specific temperatures at all. I meant that they did not look at it in the amount the probability of the specific warming would make sensible.
Is your critique that we used “severly neglected”, but you would have been ok with “neglected”? Or is your model that the scientific community does the right amount of research for different temperatures, given the likelyhood of reaching these temperatures?
Thanks for sharing this paper. I disagree with most of the key claims you make here.
Firstly, you cite the recent updates on climate sensitivity, but without clarifying that this has narrowed uncertainty at both the left and right of the tail. “The most recent estimates of equilibrium climate sensitivity show a similar distribution, narrowing the range of outcomes to exclude rises below 2 °C but not ruling out warming above 4.5 °C (Sherwood et al 2020).” The first thing to say is that equilibrium climate sensitivity measures the warming conditional on an increasing in co2 concentrations—it doesn’t tell us the unconditional probability of a particular level of warming. Moreover, this is misleading as a description of the Sherwood et al (2020) paper because it makes it sound like the right tail hasn’t changed. The paper also excludes ECS above 5.5 degrees.
Secondly, you say “GHG concentrations of 700 ppm could produce a 10% chance of exceeding a temperature rise of 6 °C (Wagner and Weitzman 2015). Such concentrations would be passed by 2100 under six of the nine of the CMIP6 SSP-RF baseline and 6.0 W/m2 forcing scenarios (Riahi et al 2017, Gidden et al 2019).” As Johannes and I discuss in our recent post, this risk now seems a lot lower
Thirdly, this is not a criticism, but the Raftery estimates have since been updated by Liu and Raftery who find “The median forecast for 2100 is 2.8 °C, with likely range (90% prediction interval) [2.1, 3.9] °C. The median is 0.4 °C lower than that of Raftery et al.1, the upper bound is 1.0 °C lower, while the lower bound is 0.1 °C higher. The tighter interval reflects the additional 5 years of data and the improved model.”
Fourthly, I’m not sure whether I have misunderstood your paper, but the entire impacts literature focuses on the impact of >3 degrees of warming. The reason for this is that it focuses on the impacts of RCP8.5 at 2100, which implies warming of 4-5 degrees. It seems like in your search strategy, you have searched for mentions of particular levels of warming. But if you had searched for mentions of RCP8.5, you would have found that almost the entire impacts literature focuses on comparing that scenario to RCP2.6 or lower.
Thank you for your feedback!
It is true that our phrasing around Sherwood et al (2020) can sound a bit misleading. However, this was not in bad faith. We did not intentionally leave out the decreased tail risks.
Overall, your criticism mainly seems to be the fact that current estimates of climate sensitivity have a smaller range that the one we used. This is true. It does not change the point of our paper though. While the range decreased, the mean basically stayed the same and if you look at our figures, this still means that there is likely to few research for the higher temperatures.
To your last point, I wasn’t aware that the impact literature mainly compares those two scenarios. However, as we look at all parts of the IPCC reports this seems to get drowned out by the other parts of the reports. We are currently doing some new analyses that try to look into this in a bit more depth.
Hello, thanks for your response.
On the Sherwood et al thing, another issue is that I don’t understand why you would use the Weitzman estimate of ECS when you have already mentioned the Sherwood et al (2020) estimate. The Sherwood et al (2020) estimate is superior, and that is clear from reading their paper. The IPCC also now accepts that it is superior. So, I don’t understand why you would mention “the most recent estimate” and then use one from 6 years ago instead.
“Overall, your criticism mainly seems to be the fact that current estimates of climate sensitivity have a smaller range that the one we used.” That is not my criticism. My criticism is that emissions look set to lower and the right tail of climate sensitivity has thinned which makes a clear material difference to your conclusions.
On your last point: this does make it look like the central claim of your paper is wrong. I am going to paste some charts from the last impacts report below on some key climate impact areas to illustrate:
Sea level
Food yields
Floods
Species loss
Fire frequency
Ocean net primary production
Precipitation
Maximum fish catch potential (A1B is between RCP6 and RCP8.5)
On drought they say “Climate change is likely to increase the frequency of meteorological droughts (less rainfall) and agricultural droughts (less soil moisture) in presently dry regions by the end of the 21st century under the RCP8.5 scenario (medium confidence). {WGI AR5 Chapter 12} This is likely to increase the frequency of short hydrological droughts (less surface water and groundwater) in these regions (medium evidence, medium agreement).”
Ocean acidification
The Sherwood reference was only included during the review process, as it was not yet published when we originally came up with the analysis. As you probably know, going from an idea to a published paper can take quite some time and you cannot read and update on all the papers that are published during that time.
I would agree that today the picture looks better as in comparison when we started working on that paper. However, predicted temperatures and mentions in the IPCC still don’t overlap and therefore we still have a research gap, albeit a smaller one.
I don’t see how this makes the paper wrong. 1) You are only referring to impact literature. It’s great if we know what impact climate change will have, but we still have a problem if we do not know how to mitigate it at higher temperatures. 2) Comparing emission scenarios is good, but every scenario has a wide range of possible temperatures. Therefore, it is also important to look at specific temperatures.
Ok, that makes sense on the Sherwood thing.
I’m not sure I understand point (1) in your last paragraph. ‘Mitigation’ with respect to climate usually refers to reducing CO2 emissions, so I don’t see how there could be a specific problem of ‘mitigating at higher temperatures’. Perhaps you mean adapting to higher temperatures?
I don’t think your point (2) hits home. The impacts literature that I have outlined and pasted above does look at specific temperatures. It looks at the most likely level of warming on RCP8.5.
In your paper you say that warming of >3 degrees is ‘severely neglected’. This is not true, and hasn’t been for years. If you had said >5 degrees that would have been true, but the claim is about >3 degrees.
That’s what I meant, sorry if I phrased this incorrectly.
I did not mean to say that they did not look at specific temperatures at all. I meant that they did not look at it in the amount the probability of the specific warming would make sensible.
Is your critique that we used “severly neglected”, but you would have been ok with “neglected”? Or is your model that the scientific community does the right amount of research for different temperatures, given the likelyhood of reaching these temperatures?