Guest Post by Willis Eschenbach
Well, I decided to take a shot at publishing my views on the cloud feedback response to increases in surface warming. I wrote it up and sent it for peer review to the Journal of Climate.
The reviewers said that it seemed like I was looking at changes in location, not changes over time. So I re-wrote it and sent it back in.
They wrote back and said ok, changes helped, and oh, by the way …
… it’ll cost you $1,546 to get it published.
I can assure you that I harnessed the awesome power that comes from splitting the infinitive. In a far-too-loud voice, I uttered various speculations regarding the ancestry, sexual habits, and personal hygiene of the Owners, Editors, and Reviewers. I fear I went so far as to encourage them to engage in auto-fornication … for all of which I’m truly sorry. It’s just what goes on in 2023, and I’m still not used to it.
So instead, I figured I’d start by publishing it here, and invite people to suggest changes, to point out inconsistencies, and to generally be some combination of Editors and Reviewers of the paper. Please be kind in your comments, I’m just a fool whose intentions are good.
With that as a prologue, here’s the current state of the paper.
Independent Climate Researcher, Occidental, California
Corresponding author: Willis Eschenbach
Cloud radiative response to a change in surface temperatures is a key component in accurately estimating future temperature changes. Changes in surface temperature lead to different cloud responses in different parts of the planet. However, the overall effect of these changes has been very poorly constrained. (Boucher 2013) Using data from satellite observations, here I develop two independent methods to estimate how the clouds in different areas respond to a surface temperature increase. Both methods show a global net surface cloud radiative cooling effect. The size of the cooling obtained in this manner is a minimum value of total cloud cooling, because more cloud-related cooling occurs as a result of a temperature-related increase in thermally-driven tropical and extra-tropical thunderstorms which cool the surface in a variety of non-radiative ways. In addition, using theoretical arguments, I show that it is unlikely that the cloud response amplifies global warming.
Clouds have a central role in modulating the global energy balance. They have long been recognized as being the major source of uncertainty in climate projections. Although a variety of evidence has been presented, a narrow constraint on how clouds respond to projected warming has remained elusive. Indeed, there is still no widespread agreement even on the sign of the cloud response to warming. Part of the challenge is that net cloud radiative feedback involves cloud effects on both solar (shortwave [SW]) and terrestrial (longwave [LW]) radiative fluxes. (Ceppi et al. 2017, Gettleman and Sherwood 2016)
2. Theoretical Arguments
A most unusual yet generally unremarked feature of the climate system is its amazing stability. Here is the maximum temperature range over a 22-year period for each 1° latitude by 1° longitude gridcell.
Fig. 1. Maximum variations in monthly temperature (trough to peak) during the period Mar 2000 – Feb 2022.
Here, we see temperature swings of over 30°C in the poles, 29°C over the land, 9°C over the oceans, and 14.8 °C for the globe as a whole. But despite those large intra-annual swings, after 12 months the temperature always returns to nearly the same value. Over the same Mar 2000 – Feb 2022 period, the CERES data reveals annual average global surface temperature changes of only about 0.5°C, which is only three percent of the intra-annual variation.
Figure 2. Annual temperature ranges for the globe (red line) and monthly temperature ranges for parts of the globe. Red line shows range of 20th Century global average annual temperatures, around 0.3%.
As Figure 2 above shows, this amazing longer-term stability cannot be from thermal inertia, given the far larger monthly intra-annual swings. This overall steady-state condition argues strongly for the existence of natural thermoregulatory phenomena opposing any change in the overall steady-state temperature.
This is supported by Le Chatelier’s Principle. Le Chatelier enunciated a simple principle that governs systems that are in a steady-state condition. Le Chatelier’s principle asserts that a disturbance applied to a system at a steady state may drive the system away from its equilibrium state, but will invoke a countervailing influence that will counteract the effect of the disturbance. (Gorshkov et al. 1990) This principle strongly suggests that if the global average temperature changes, the clouds and other phenomena will act to counteract the temperature change, not to reinforce it.
3. Observational Data Analysis
The net cloud radiative effect (CRE) at the surface is composed of the clouds’ effect on two different types of radiation. The first is solar (shortwave) radiation, which is both reflected and absorbed by clouds. The second is thermal (longwave) radiation, which is both radiated and absorbed by clouds. The net surface CRE, which I’ll call “CRE” for simplicity, is the total of the two effects at the surface where we live. In general, clouds cool the surface. Figure 2 shows the global variations in the CRE. In Fig. 2 we see that the clouds warm the poles and the deserts, and cool everywhere else.
Fig. 3. Surface cloud radiative effect, on a 1° latitude x 1° longitude basis.
The short-term change in surface CRE with temperature is easily calculated using the CERES data. Figure 4 shows that result.
Figure 4. Short-term trends in surface cloud radiative effect as a function of temperature. Trends are ordinary least squares linear regression slopes.
However, that doesn’t tell us what we need to know, which is how the clouds respond to a long-term change in surface temperature. Despite that, there are two ways that we can use observational data to measure that response.
Both of them depend on a simple idea—as a long-term average for each gridcell, over thousands of years, the temperature and the corresponding cloud radiative effect have reached a steady state condition. All of the various phenomena affecting the CRE, such as relative humidity, boundary-layer inversion strength, CAPE (Convective Available Potential Energy), oceanic subsidence and upwelling, and other factors now oscillate around long-term average values for each given gridcell. Thus, the average relationship between temperature and CRE for each gridcell represents the long-term steady-state relationship.
The first way to see what will happen if the surface temperature warms is a gridcell-based scatterplot of CRE and temperature.
Fig. 5. Scatterplot, 22-year averages of CRE versus surface temperature. Each dot is a 1° latitude by 1° longitude gridcell.
Despite this scatterplot including both land and ocean and covering from the topics to both poles, there is a clear pattern. Looking from the left to the right in the scatterplot, the slope of the black/white line shows the direction and amount of change in the CRE as the temperature increases. There are four different zones.
The coldest zone encompasses the Antarctic and Greenland ice caps. Where the average monthly gridcell temperature is below -20°C, you are in one of those two locations. There, increasing temperatures lead to increasing cloud warming. This represents less than 4% of the planetary surface.
The next zone is from -20°C to 10-15°C. In this zone, increasing warming results in increasing cloud cooling. The third zone is from 10-15°C to about 25°C. In this zone, increasing temperature leads to increasing cloud warming.
Finally, in the warmest parts, increased surface warming leads to greatly increased cloud cooling. At its greatest, an increase of 1°C leads to up to 40 W/m2 of increased cloud cooling (reduction in downwelling surface radiation).
Now, this shows us the overall pattern of the relationship between temperature and CRE. It is extremely non-linear. But it is a general indication, with lots of scatter around the trend line. It also shows areas from all around the world combined.
What this method doesn’t show is either the detailed spatial pattern or the area-weighted global average response of the CRE to increasing temperature. For this I use a second method.
The second method looks only at the average values of the gridcells in the area immediately around each gridcell. Consider a gridcell in the ocean as an example. Nearby gridcells to the north, south, east, and west of that chosen gridcell will have different long-time average values for temperature and CRE. So we can determine the long-term effect by looking at the local relationship between average temperature and average CRE. For each gridcell, I have used a box that is 9° latitude by 9° longitude, centered on the chosen gridcell, and I’ve used a linear regression of that block of data to determine the value for the center gridcell. I’ve analyzed the land and the ocean separately, to avoid mixing different regimes. However, this seems to make little difference. The result is shown in the following Pacific- and Atlantic-centered graphics.
Fig. 6. Changes in surface cloud radiative effect per 1°C change in surface warming. The lower panel is the same as the upper but with an Atlantic-centered view. All values used in the calculation are the average of the full 22 years of the CERES record.
This shows two views, one Pacific and one Atlantic centered, of the detailed location and size of changes in CRE from 1°C surface warming. Globally, there is an area-averaged net cooling of -1.7 W/m2. The main cooling occurs over the ocean, with an area-averaged cooling of -2.4 W/m2. The land is the only area which is even slightly positive, with an area-averaged warming of +0.3 W/m2
These results are in good agreement with those of Ramanathan and Collins (Ramanathan, V., & Collins, W. (1991)), although the proposed mechanisms are different, and these results are for the planet while Ramanathan and Collins only looked at the Pacific Warm Pool.
3. Stability And Uncertainty
If this metric is indeed a measure of the long-term change in CRE with warming, it should change very little from year to year. The boxplot below shows 22 CRE feedback values for each geographical area listed in Figure 6, one for each year of the CERES record.
Figure 7. Boxplot, change in CRE from 1°C surface warming. Data for each of the 22 years in the CERES record.
As expected, there is very little variation in the results despite the shortness (one year) of each dataset. This indicates that even a 22-year average will give accurate values for the change in surface CRE per 1°C warming. As in Figure 6, the only large area that shows positive feedback is the land, and the feedback is quite small.
4. Data Details
I used monthly gridded Clouds and the Earth’s Radiant Energy System (CERES) Energy Balanced and Filled Edition 4.1 data (Loeb et al. 2018). The CERES record is quite stable (Loeb et al. 2016), which makes it an excellent record for this type of analysis. All of the CERES data used covers the 22-year period from March 2000 to February 2022.
For surface temperature, I have used the CERES surface upwelling longwave dataset, converted to temperature using the Stefan-Boltzmann equation. For verification of the calculated CERES surface temperature data, I have compared it with the results using the Berkeley Earth gridded land/ocean data record (Rohde and Hausfather 2020). The area-weighted average difference between the two is only 0.43°C. This difference is not surprising because the Berkeley Earth dataset is a combination of air temperature over land and sea surface temperature. On the other hand, the CERES data is surface temperature everywhere. Below is the same calculation shown in Figure 5 but using the Mar 2000 – Feb 2022 Berkeley Earth Data in place of CERES data for that same period. Note that there is very little difference between this and Figure 5 above which uses CERES data.
Figure 8. As in Figure 3, but using Berkeley Earth surface temperature data in place of CERES data.
5. Final Thoughts
As mentioned above, the cloud radiative effect is only one of the ways that the clouds affect the surface temperature. In addition, thunderstorms cool the surface by means of:
• Increased surface albedo over the ocean due to the white surface foam, spume (foam driven aloft by the wind), and spray.
• Increased evaporative cooling due to the thunderstorm-generated winds at the base, as well as from the provision of dry air to the surface.
• Increased evaporation from the increase in surface area on the millions of spray droplets.
• Cold wind and rain directly cool the surface.
• Increased albedo due to the tall cumulonimbus towers, particularly in the afternoons.
• Increased radiation to space due to the lack of water vapor in the dry descending air between the thunderstorms.
Tropical thermally driven thunderstorms increase with increasing temperatures. As a result, the cloud radiative cooling (CRE) is enhanced by increased thunderstorm production, and the CRE cooling estimates represent a minimum value.
All of this work is my own. However, I owe immense thanks to all of the outstanding scientists who have preceded me. I have no conflicts of interest.
Data Availability Statement.
The underlying CERES EBAF 4.1 data is NASA/LARC/SD/ASDC, 2022. CERES Energy Balanced and Filled (EBAF) TOA and Surface Monthly means data in netCDF Edition 4.1., accessed 11 December 2022, https://ceres.larc.nasa.gov/data/#energy-balanced-and-filled-ebaf.
The underlying Berkeley Earth data is Berkeley Earth, 2022, Monthly Land + Ocean Average Temperature with Air Temperatures at Sea Ice, accessed 17 December 2022, https://berkeley-earth-temperature.s3.us-west-1.amazonaws.com/Global/Gridded/Land_and_Ocean_LatLong1.nc
Boucher, O. et al., 2013: Clouds and aerosols, Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (Cambridge University Press, Cambridge, UK, 2013), pp. 571–657.
Ceppi, P, F. Brient, M. D. Zelinka, D. L. Hartmann, 2017: Cloud feedback mechanisms and their representation in global climate models. Wiley Interdisc. Rev. : Clim. Change 8, e465.
Gettelman, A., Sherwood, S.C., 2016: Processes Responsible for Cloud Feedback. Curr Clim Change Rep 2, 179–189. https://doi.org/10.1007/s40641-016-0052-8
Gorshkov, V.G., Sherman, S.G. & Kondratyev, K.Y., 1990: The global carbon cycle change: Le Chatelier principle in the response of biota to changing CO2 concentration in the atmosphere. Il Nuovo Cimento C 13, 801–816 https://doi.org/10.1007/BF02511997
Loeb, N. G. et al., 2018: Clouds and the Earth’s Radiant Energy System (CERES) Energy Balanced and Filled (EBAF) Top-of-Atmosphere (TOA) Edition-4.0 data product. J. Clim. 31, 895–918.
Loeb, N., N. Manalo-Smith, W. Su, M. Shankar, S. Thomas, 2016: CERES top-of-atmosphere Earth radiation budget climate data record: Accounting for in-orbit changes in instrument calibration. Rem. Sens. 8, 182.
Ramanathan, V., & Collins, W. (1991). Thermodynamic regulation of ocean warming by cirrus clouds deduced from observations of the 1987 El Niño. Nature, 351(6321), 27–32. doi:10.1038/351027a0 https://sci-hub.se/10.1038/351027a0
Rohde, R. A. and Hausfather, Z., 2020: The Berkeley Earth Land/Ocean Temperature Record, Earth Syst. Sci. Data, 12, 3469–3479, https://doi.org/10.5194/essd-12-3469-2020.
So, there it is. All comments, criticisms, and improvements gladly considered. This site provides one of the best peer-review processes on the planet.
Please remember that when you comment, you need to quote the exact words you are discussing. That way, we can all be clear about your subject.
Finally, a couple of requests.
First, I’d still like to get this sucker published. Does anyone know of a reasonably high-impact-factor journal that does NOT charge $1,546 to publish a paper?
And second, does anyone want to take this paper over and shepherd it through the review process? This method worked out very well for Craig Loehle and me. With some assistance from me, he rewrote parts of my post on extinctions entitled Where are the Corpses?, he arm-wrestled the journals, and we got it published as Historical bird and terrestrial mammal extinction rates and causes, with him as the Lead Author. So far, about 150 citations, with more each month.
Anyone interested? Any “publish or perish” folks out there not interested in perishing? Because I truly hate dealing with the journals …
My very best wishes to all, enjoy this marvelous world where we are given so little time.