Hansen on Global Warming in the Pipeline
A paper is currently sitting on the ArXiv[1] awaiting publication that has garnered a fair amount of attention on social media. It's a paper by Hansen with several other well-respected and influential scientists. Since this paper is currently in prepublication, and I don't know what this paper will be like when it's published, I don't want to make too much of this, but it is very interesting, if alarming. At best, I see this as an upper bound estimate of how bad AGW could become long-term. I'd like to consider what this paper (in it's current form) is suggesting and evaluate what it claims.
Summary of Hansen's Paper
The general thrust of this paper is that paleoclimate evidence shows that fast feedback sensitivity (ECS) is 3.5 - 5.5°C and GHG forcings are are 4.1 W/m^2. After slow feedbacks bring the Earth's climate system into full equilibrium with these forcings (what is called Earth System Sensitivity), we can expect about 7-10°C warming long term from the forcings that human activity has already put in the pipeline. These statements are very alarming, but there are two recent developments in paleoclimate studies that Hansen et al believe warrant these conclusions.
First, warming since the LGM is more than previously estimated, implying that ECS is higher than 3°C. Earlier work from Shakun and others have placed the difference between the LGM and HTM to be about 4°C. However, a more recent reanalysis (that has the benefit of resolving some seasonal biases) estimate this difference to be 6-7°C. Since the forcings are the same (they estimate 3-5 W/m^2), the larger increase in temperature implies a larger sensitivity. So if a 3-5 W/m^2 increase causes 6°C warming, then λ must be ~1.2 - 2°C/W/m^2. That implies a sensitivity of at least 4°C for 2xCO2. "We conclude ECS is at least approximately 4°C and is almost surely in the range 3.5-5.5°C. The IPCC AR6 conclusion that 3°C is the best estimate for ECS is inconsistent with paleoclimate data."
Second, aerosol pollution is on the decline, so the masking effect of CO2 is being diminished. Human aerosol pollution has masked some of the warming from our greenhouse gas emissions. But as we continue to clean up our air, aerosol pollution is likely to continue to decrease, and the masking effect of aerosols will shrink, so future warming will follow a track that matches the increase in greenhouse gases more closely.
Hansen et al's concern here is not so much with projections through 2100 but with long-term warming in the "pipeline" from current conditions on time scales taking millennia to be realized. ECS estimates incorporate only rapid feedbacks, and it's a useful metric for the short-term effects of increases in GHG concentrations. But over longer time scales, the Earth's climate system continues to respond to this warming signal through the poleward movement of boreal forests and the shrinking of terrestrial ice sheets. These decrease the Earth's albedo, so less solar energy is reflected and more is absorbed by land and oceans, amplifying warming. These long-term changes are part of Earth System Sensitivity (ESS). As a general rule of thumb, scientists have seen ESS to be proportional to ECS, with a more or less constant ESS/ECS ratio. Lund et al estimated this ratio to be between 1.4 and 1.5[2] Yet Hansen's estimate is significantly higher. "When all feedbacks, including ice sheets, are allowed to respond to the climate forcing, the equilibrium response is approximately doubled, i.e., ESS is ~ 10°C." So for Hansen et al, ESS/ECS ≈ 2.5. Without aerosol pollution, Hansen et al conclude "equilibrium warming for today’s GHG level, including slow feedbacks, is about 10°C. Today’s level of particulate air pollution reduces equilibrium warming to about 7°C."
 |
| Hansen's Long-Term Temperature Projections from Current Forcings |
What Hansen et al appear to be arguing is that recent developments in paleoclimate studies show that we have been underestimating both ECS and ESS. Consequently there is a lot more warming in the pipeline on millennial time scales than has been previously estimated. At full equilibrium with current forcings, we can expect "7-10°C global warming" above preindustrial levels. This is illustrated quite nicely in Figure 7 above. The blue line is actual global temperatures. The red/green (?) line following 1750 shows the estimated increase in radiative forcings due to anthropogenic activity from GHGs alone. The black line shows GHG + aerosol forcings; this line represents the most likely forcings from current conditions. The temperature anomaly scale on the left is scaled to the expected ESS full equilibrium temperature from those forcings. So without the making effect of aerosols, Hansen expects ~10°C warming from current forcings. With aerosols, this is reduced to 7°C, though this warming is slowly realized over thousands of years.
Response to Hansen's Paper
In 2016, Carolyn Snyder published a paper[3] making a similar estimate that ESS is large. Her estimate was 9°C (95% CI from 7 - 13°C). If ECS = 4°C, then for her ESS/ECS = 2.25. Her conclusion was that "stabilization at today’s greenhouse gas levels may already commit Earth to an eventual total warming of 5 degrees Celsius (range 3 to 7 degrees Celsius, 95 per cent credible interval) over the next few millennia as ice sheets, vegetation and atmospheric dust continue to respond to global warming." By all accounts I've read, Snyder's paper was well-received by all for its reconstruction of global temperatures for the last 2 million years, but she received significant criticism for estimating ESS/ECS to be too large. Most estimates I've seen estimate ESS ≈ 5°C or an ESS/ECS ≈ 1.5. The initial response to Hansen's prepublished paper appears to be similar. I've seen responses from Mann[4], Gavin Schmidt[5], and Ken Rice[6] suggesting that this paper overestimates the amount of warming in the pipeline from current forcings.For myself, I have been intrigued by Osman's reconstruction[7] which was influenced by Tierney's work[8] on the amount of warming from the LGM to the HTM. The increase in estimated warming should in principle lead to an increase in estimates of ECS from this paleoclimate data, provided that the forcings estimates remain roughly the same. The work of Hansen et al and Snyder do deserved to be taken seriously. However, the more standard estimates for ECS and ESS are there for good reason, and I'm not one to jump on a high sensitivity bandwagon on the results of just a couple studies, when the weight of evidence we have for ESS and ECS suggest that they're probably lower. It seems to me much of Hansen et al's estimate of ECS depends on there being 6-7°C warming from the LGM to the HTM, but this estimate is far less certain than estimates made from the instrumental record. And while this papers takes forcings to be 3-5 W/m^2, you have to account for all forcings to get at sensitivity (λ). Efforts to account for all forcings have yielded results as high as 6.5 W/m^2 (this is the most common that I've seen). If this is correct, then 6°C warming with 6.5 W/m^2 is λ = 6/6.5 = 0.92°C/W/m^2 or ECS = 3.4°C. And for myself, using data from HadCRUT5 and CO2 concentrations with EEI estimates, I find ECS = 3.2°C, and that is very consistent with the IPCC's own estimates and range. And while my own calculations are admittedly simple, I don't see compelling reasons to think that the consensus reports from the IPCC are wrong. Is it possible that Hansen's estimate of ECS = 4°C is correct? Certainly. But it still seems to me more likely that it's closer to 3°C.
More important here, though, I think, is that the warming from the LGM to the HTM occurred for different reasons than current warming. Hansen sees ESS as basically a sum of GHG forcings and ice sheet forcings with feedbacks, and I think he assumes that the ice forcings are dependent on changes in GHG concentrations. This is how I understand him saying, "the improved ΔT to Fghg scale factor in Fig. 7 is (Fghg + Fice)/ Fghg × 1°C per W/m2 = 2.4°C per W/m2 of GHG forcing." Schmidt seems to be taking issue with this, and as best I can tell, I agree entirely with Schmidt. I don't think this is right. The glacial cycles of the Quaternary are synced to orbital forcings; initially with obliquity (41K years) but then after the mid-Pleistocene transition (MPT) with eccentricity (100K years). These forcings are small globally, since they do not change the amount of energy received by the Sun over the course of a year. They change the regional and seasonal distribution of energy the Earth receives within a year. In effect these orbital forcings periodically shine a heat lamp on one of the poles, while other parts of the Earth are correspondingly cooler. Regional warming begins to melt ice sheets, albedo decreases, and greenhouse gas concentrations increase as the oceans degas CO2 as a feedback that amplifies warming. Other forcings such as dust also play a role. But the percent of land covered by ice sheets decreases significantly more from the LGM to the HTM than it can from current warming moving forward; much of the terrestrial ice sheets have already melted. So there may be unique features of these glacial cycles that make them poor analogs to current warming and that makes Hansen's equation less legitimate for projections from current conditions, which are forced primarily by increases in GHGs. So it may be that Hansen's estimate of ESS is biased by the unique conditions affecting warming out of the LGM. That's my take; I could be wrong.
I think a better way to estimate ESS would be to look at analogs to current forcings, and the best analog is mid-Pliocene climate, the last time CO2 levels were ~400 ppm. Global temperatures during this time frame averaged 2 - 4°C warmer than preindustrial levels. This is more consistent with ECS = 3°C and ESS/ECS = 1.5. Now it's certainly possible that warming from other GHGs like N2O imply that full equilibrium to current conditions would exceed 4°C, but I think ~7°C is a bit of a stretch. At the very least, I think Hansen et al should show how their results are consistent with mid-Pliocene climate. I'm hoping that the final version of this paper does this. As I see it, the constraints on ESS and ESS/ECS from the mid-Pliocene make Hansen's estimates less likely.
That said, I think it valid to consider the impacts of this paper should it be 100% correct. If ECS = 4°C and ESS/ECS = 2.5, how should we expect this to impact us? My first thought is that the time scale is long enough that it does not really increase the immediacy to the problem. Short-term through 2100, expected warming is probably not too much different from the IPCC's projections. And over the next few hundred years, I think we would certainly be able to develop technologies to help mitigate these long-term impacts. It's the probable effects of our forcings on centennial time scales that produce a much stronger argument for mitigation over the next couple decades.
My second thought is that there are two types of sensitivities that are relevant here. The first is the sensitivity of temperature to forcings, primarily GHGs. We have different metrics for this depending on time scales (TCR, ECS, and ESS), but we are concerned here with how increases in CO2 impact global surface temperature. But there's a second kind of sensitivity here, and that is the sensitivity of planet to temperature, say, the temperature differences between an Earth with no ice sheets, our current Quaternary interglacial, a Quaternary glacial maximum, and a "snowball earth" episode. Maybe we could call this Earth-Temperature Sensitivity (ETS), and it could be expressed as perhaps a change in the percent of the Earth covered by ice sheets per unit change in global surface temperature. This kind of metric would only be useful at the ΔT from the warmest "snowball earth" temperature and the coolest "no ice sheet" temperature, and I would be surprised if it were linear. But what strikes me as important is that, regardless of the ΔT between the LGM and the HTM, changes in amount of terrestrial land ice, sea levels, and the planet's vegetation is pretty well-known. If the ΔT is closer to Shakun's estimate of 4°C, then ECS and ESS/ECS are lower, but the changes in the Earth are greater for each 1°C increase in temperature (ETS is high). However, if Osman's estimate of 6-7°C is correct, ECS and ESS/ECS are higher, but the changes in the Earth are correspondingly less for each 1°C increase in temperature (ETS is low). Projecting forward, then, Hansen's projections may be for more warming from current forcings, but with a correspondingly smaller amount of impact for each 1°C of warming.
Conclusion
I see this paper as an upper-bound, worst case estimate of the long-term effects of human activity on global temperatures. While possible, I don't see these high sensitivity estimates as being more plausible than those from the IPCC. That said, I could be wrong, and the warming from current forcings could be high, but I also fully expect future technologies will be able to mitigate those impacts. The expected impacts through 2100 still require our immediate mitigation efforts to avoid the worst impacts of AGW that we should expect to realize on much shorter time scales.
References:
[1] Hansen et al. Global Warming in the Pipeline. ArXiv.
https://arxiv.org/ftp/arxiv/papers/2212/2212.04474.pdf
[2] D.J. Lunt, A.M. Haywood, G.A. Schmidt, U. Salzmann, P.J. Valdes, and H.J. Dowsett, "Earth system sensitivity inferred from Pliocene modelling and data", Nature Geoscience, vol. 3, pp. 60-64, 2009. http://dx.doi.org/10.1038/NGEO706
[3] Snyder, C. W. (2016). Evolution of global temperature over the past two million years. Nature, 538(7624), 226–228. doi:10.1038/nature19798
[4] Michael Mann. https://twitter.com/MichaelEMann/status/1603446912073764865
[5] Gavin Schmidt. https://www.realclimate.org/index.php/archives/2016/09/why-correlations-of-co2-and-temperature-over-ice-age-cycles-dont-define-climate-sensitivity/
[6] Ken Rice. Hansen's 10 C. And Then There's Physics. https://andthentheresphysics.wordpress.com/2023/01/11/hansens-10c/
Comments
Post a Comment