No, it probably isn’t mostly due to changes in clouds!

I haven’t done a paper debunk for a while, but a reader got in touch to ask about a recent paper by Hans Rolf-Dübal and Fritz Vahrenholt, so I thought I would have a quick look. The paper is Radiative Energy Flux Variation from 2001–2020 and is in the open-access journal Atmosphere. The paper doesn’t actually draw any strong conclusions, but it does stress that [t]he declining TOA SW (out) is the major heating cause (+1.42 W/m2 from 2001 to 2020), where TOA stands for top-of-atmosphere, and SW is short-wavelength.

What the paper is essentially claiming is that most of the warming over the period 2001-2020 is due to a reduction in cloud albedo which then leads to more absorbed solar radiation. This has then been used by some to claim that [c]hange in clouds likely cause of warming in the past 20 years.

I haven’t worked through the details in the paper to really know if the results they present are correct, but I don’t think the results presented in the paper are necessarily all that surprising. I think it is mostly related to what I was highlighting in this post.

There can sometimes be a rather simplistic idea about how global warming actually happens. The simple view is that adding greenhouse gases to the atmosphere reduces the outgoing longwavelength flux, producing a planetary energy imbalance, which then causes the system to warm until it returns to energy balance. What actually happens is – unsurprisingly – a little more complicated.

It is correct that adding greenhouse gases to the atmosphere does reduce the outgoing longwavelength flux and does produce a planetary energy imbalance. However, as pointed out in this paper, as the system warms, there are then feedback responses that can further enhance the reduction in outgoing longwavelength radiation (water vapour). However, there are also other responses that can counter-act this (clouds), and others (also clouds) that can lead to a reduction in albedo and, consequently, an increase in the amount of absorbed solar radiation.

As this paper then illustrates, when energy accumulates in the climate system due to an enhancement in atmospheric greenhouse gas concentrations, this accumulation is primarily due to an increase in absorbed solar radiation, rather than simply being due to an imbalance in the long-wavelength fluxes. However, this doesn’t somehow contradict that adding greenhouse gases to the atmosphere will cause the system to warm, or suggest that changes in clouds are causing most of the warming.

So, although I haven’t work through the Dübal and Vahrenholt paper in detail, the basic result they present seems broadly consistent with what is expected. That they find that most of the warming over the 2001-2020 period was due to a reduction in cloud albedo doesn’t really contradict our understanding of greenhouse warming and doesn’t suggest that most of the warming over this period was due to changes in clouds.

Most of the warming is almost certainly due to the human emission of greenhouse gases into the atmosphere. How clouds then respond to the subsequent warming then leads to most of the accumulated energy being due to an increase in absorbed solar radiation. If anything, as highlighted in the video in this post, this might actually be suggesting that equilibrium climate sensitivity is well above 2oC, rather than highlighting some major challenge to our understanding of greenhouse warming.


Outgoing longwave radiation – post I wrote explaining why most of the accumulated energy is due to increased absorded solar radiation.
Global warming due to increasing absorbed solar radiation – paper by Trenberth and Fasullo.
Shortwave and longwave radiative contributions to global warming under increasing CO2 – paper by Donohue et al.

This entry was posted in Climate change, Climate sensitivity, Global warming, Greenhouse effect and tagged , , , , , . Bookmark the permalink.

34 Responses to No, it probably isn’t mostly due to changes in clouds!

  1. typo in second sentence of 4th paragraph: “…which then causes the system to warm until it returns to energy imbalance.” I think you meant say “returns to energy balance”

    I am not sure I am going to be able to wrap my head around this post to any significant extent, but I will read through a couple times and follow the comments if time allows.
    Thanks for your work trying to make this stuff more accessible to non-scientist types.

  2. Chubbs says:

    Yes, not surprising. I left a comment on your prior post with a link to an Isaac Held blog. Bottom-line, OLR increasing as CO2 increases, indicates positive cloud feedback and high ECS.

  3. bjchip says:

    Always regarded any alterations in the imbalance to be, ultimately, just additional forms of feedback. Ice in the Arctic and warming induced water in the air, and the question of *which* clouds is, of course, a wonderfully deep hole for the rabbits.

  4. paulski0 says:

    When I started reading I thought the post might be about this paper. Interestingly, that one is prominently cited by the paper actually being discussed here. And they suffer the same basic error: they assume the effects warming by greenhouse gases should produce without ever bothering to check what actual models produce, despite relevant model data being extremely available. Isn’t this something which any reasonable peer review should pick up on?

  5. Paul,
    You would think so, but then we all probably know that peer review isn’t perfect.

  6. paulski0 says:


    A key thing to recognise is that the Earth can only output more longwave as it warms, regardless of what’s causing the warming.

    First imagine a system where only longwave is involved. It starts at equilibrium so incoming energy matches outgoing energy. Now we increase greenhouse gases. This first causes a reduction in outgoing longwave, but no change in incoming. Then this causes warming, which increases outgoing longwave. We get back to equilibrium when enough warming has happened that outgoing longwave again matches incoming.

    Now, let’s introduce shortwave. Shortwave is incoming energy from the Sun, which is fairly constant, so what matters with shortwave is how much the Earth is reflecting back out to space without absorbing. The most obvious factors are ice and clouds. What happens with warming: melting ice. This increases the amount of shortwave absorbed by the Earth/reduces the amount immediately reflected back out to space, which causes warming. Here the direct cause of warming is in the shortwave (albeit ultimately caused by greenhouse gas increase), but refer back to my first sentence: even though the warming was caused by reduced outgoing shortwave, the only way the Earth can respond to plug the energy imbalance is by outputting more longwave.

    Another less obvious, but very important, shortwave factor is water vapour, which absorbs shortwave. With warming comes a huge increase of water vapour in the atmosphere, and that also results in decreasing outgoing shortwave. Again the Earth can only counter that introduced imbalance by outputting more longwave.

    So, given these ice and water vapour feedback factors outgoing shortwave can only really reduce, because there’s nothing to counter it. The only a priori plausible reason this might not happen is a strong negative cloud feedback, but that doesn’t seem to be a reality.

    The Earth’s compensatory mechanism – producing more outgoing longwave via warming – occurs for both shortwave and longwave factors. Transiently the net effect is somewhat complex, depending on rate of warming vs. rate of GHG increase. Historically I think it’s still likely that outgoing longwave is reduced relative to pre-industrial even though the trend is now going up. At equilibrium outgoing longwave will be substantially greater than pre-industrial because of the substantial reduction in outgoing shortwave.

  7. Nepal says:


    You make a good point that CO2-induced warming may decrease outgoing shortwave radiation, primarily through reduction of sea ice and snow. However, this paper claims to identify the SW forcing primarily to be reduced cloud albedo. I have not checked if their methodology is sound, but if so I don’t see how CO2 warming would result in reduced cloud cover. Usually the opposite is predicted.

  8. Nepal,
    As I understand it, what is expected from clouds isn’t exactly clear. This paper seems to be suggesting that most of the cloud responses are expected to produce a positive cloud feedback.

    The TED talk by Kate Marvel that I highlighted in this post is also good, but I haven’t had a chance to look at it again, so can’t quite remember what Kate concluded.

  9. Chubbs,
    Thanks for that link to Isaac Held’s post. It’s really good.

  10. Nepal,
    Actually, I have just listened to Kate Marvel’s TED talk, and at about 10:35 Kate starts talking about observations suggesting that some clouds are moving from the tropics towards higher latitudes. This would, I think, act to increase absorbed solar radiation.

  11. Ben McMillan says:

    I guess as usual it is worth pointing out the Journal ‘atmosphere’ is another MDPI thingy: the article was published 5 weeks after the journal received it, like many others in this issue, so you wonder how much of the usual peer-review process took place.

  12. Chubbs says:

    Nepal is mistaken. GHG warming reduces low cloud amounts on average. The effects depend on cloud type and location. The shifting north mentioned above has been found to be less important than reduction in low cloud amount. There is a good discussion in the recent IPCC work group 1 report Chapter 7 below (p 64-69)

    Click to access IPCC_AR6_WGI_Chapter_07.pdf

  13. Chubbs, thanks I hadn’t seen that.

  14. angech says:

    Change in cloud cover is very important in the El Niño / La Niña cycle.
    The BOM dedicates a graph to it
    These changes do seem to coordinate with extra warming or cooling of the pacific oceans and the global temperature.
    Some scientists have placed a lot of emphasis on cloud changes as the main mechanism.
    This is not generally accepted
    One of the problems is the overall humidity of the globe is pretty invariant which suggests that cloud cover might be as well.
    On the other hand winds precipitation and land structures like mountains and lakes plus seasonal changes mean that clouds may be more important than we realise

  15. angech,
    Noone is disputing that cloud changes are important. The main point is simply that they’re a feedback that can amplify warming, not an independent driver of warming. They clearly are important, but that doesn’t still doesn’t mean that changes in cloud caused the warming.

  16. Nepal says:

    Thank you ATTP, Chubbs.

  17. Chubbs says:

    The root problem here is that skeptics are unaware of science. The radiation changes that satellites are observing is exactly what climate models predict. Of course being unaware, wrong even, never holds then back.

  18. Tony Banton says:

    There is this paper that suggests that it’s the -ve phase of the PDO wot done it. …..

    “Earth’s Albedo 1998–2017 as Measured From Earthshine”

    “ The recent drop in albedo is attributed to a warming of the eastern pacific, which is measured to reduce low-lying cloud cover and, thereby, the albedo.”

  19. Tony,
    I thought I’d seen some criticisms of that paper, but maybe I’m thinking of another one. If I do find it, I’ll post it here. [Edit: I must be thinking of another paper, as this one is pretty recent.]

  20. I’m thinking of this. I think some of the early Earthshine papers were suggesting that global warming could be due to changes in clouds. Even though some of the authors of the paper that Tony highlighted were on those early papers, I think this new one is actually quite reasonable.

  21. russellseitz says:

    The to and fro on IR radiative forcing versus aerosol and cloud albedo is about to turn 100.

    A reasonably disinterested observer reviewing the state of the climate debate in 1924 for the Carbon volume of his well and truly Comprehensive Treatise On Inorganic Chemistry noted that while the case for CO2 warming as made by everyone from Tyndall to Arrhenius seemed sound to him there were none the less some meteorologists who maintained volcanic albedo modulation was the prime climate forcing, and who accordingly published up a storm in the decade following the 1910 Katmai eruptions

    A century later we know who won that one:

    As a matter of mere AGU meeting structural anthropology , something similar happened when the KT impact hypothesis was floated — contrarian vulcanologists closed ranks against Alvarez, insisting the dinosaurs must have been cooked and the sun blotted out by K-T era flood basalt eruptions . If you have a hammer…..

  22. angech says:

    “angech says:
    November 28, 2021 at 12:36 pm
    Change in cloud cover is very important in the El Niño / La Niña cycle.”

    So? It’s a cycle and there are no lingering effects.

  23. Chubbs says:

    Isn’t earthshine the same as TOA SW flux? If so, the earthshine result is in agreement with the satellite radiation measurements and the cause is greenhouse warming.

  24. Chubbs,
    Yes, I think so. However, I think some of the early Earthshine papers were suggesting that maybe it was an indication of a different cause, but that was probably due to a misunderstanding of how greenhouse warming would then impact the LW and SW fluxes.

  25. JCH,
    Yes, there’s a link to that paper at the end of the post.

  26. JCH says:

    Didn’t see that. Sorry.

    If the rate at which energy is lost to space is not slowed, the GHE, then additional absorbed solar is going to leave the system at the same rate as it did preindustrial. How can that not be a big problem for their notion?

  27. Chubbs says:

    JCH, Yes if it was only clouds, outgoing IR would increase. There would also be changes in the patterns of warming. For instance would expect biggest warming during daytime instead of night and at lower latitudes instead of arctic; and the stratosphere would warm instead of cooling. Maybe other changes as well

  28. russellseitz says:

    Chubbs, it seems to take several iterations per generation to erase the confusion that arises from wavelength and temperature dependent radiative forcing feedbacks. The first such iteration took place about a century ago , and led to a generation of debate about CO2 doubling sensitivity . A debate that, as today, pitted anecdotal meteorological evidence against physical calculations.

    Absent computers ,atmospheric spectroscopy informed by quantum mechanics, and satellite observations it did not get very far or advance very fast:

    Then as now, there were those who extended the controversy by refusing to acknowledge the uncontroversial.

  29. Mal Adapted says:

    A debate that, as today, pitted anecdotal meteorological evidence against physical calculations…

    Thanks Russell, I hadn’t known about that episode in the history of climate science. My reaction to Abbott and Fowle was the same as yours. AGW denial hasn’t gotten any more sophisticated in a century, has it?

  30. Joshua says:

    Interesting thread.

    Kudos to Judith for crediting science that runs counter to her typical narrative.

    And Tol obnoxiously goes the ad hom and guilt by association routes (instead of judging science on its scientific merits) as indeed we might well have predicted:

  31. Tol thinks he’s in a position to criticise others for making a fool of themselves 😂

  32. Dave_Geologist says:

    And of course the first comment on that PNAS Twitter thread was a brain-dead one.

    Plus ça change.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.