Head in a Cloud

troposphere and stratosphere meet blogosphere

Head in a Cloud random header image

GCR’s, Clouds, and Climate, Oh My!

July 31st, 2006 by Sean Davis · 13 Comments

If you were at a BBQ with a group of atmospheric scientists, and wanted to raise some hackles, there are several subjects that are notoriously easy to get people riled up… …Talk of Hockey Sticks, Inhofe, Crichton, Coulter, and you’re sure to get people going… …Bring up the effects of galactic cosmic ray’s (GCR’s) on climate, and you’re sure to lure some group of otherwise peaceful scientists into a gloves-off throwdown!

I haven’t seen any papers on this subject in a while, but just this week in GRL a new paper was published by Luis Eduardo Antunes Vieira and Ligia Alves da Silva of the Instituto Nacional de Pesquisas Espaciais, Brazil, entitled “Geomagnetic modulation of clouds effects in the Southern Hemisphere Magnetic Anomaly through lower atmosphere cosmic ray effects”.

As a bit of background, the basic ideas behind the potential GCR-climate connection is that GCR’s increase low cloud cover via increased CCN production (via increased atmospheric ionization), which acts as a cooling effect on the climate. There are several reasons that the GCR-climate connection has been controversial. The good folks over at Realclimate.org have several posts on the topic (here, here, and here), and I don’t really feel jumping into the fray and taking sides on this issue.

My impression of this topic is that some (on both sides of the “climate war”) are using it to wage a proxy battle over climate policy. From the AGW contrarian standpoint, the existence of a GCR-climate connection could “prove” that the global warming is of solar, and not anthropogenic, origin, and therefore mandate that nothing be done about GHG emissions. I’ve oversimplified the logic of course, but one could also imagine a desire to disprove a GCR-climate connection as a means to promote action on climate change. Again, this is my impression, and I don’t want this blog to be another proxy ground for the climate wars. …Readers interested in the climate wars would be much better off following one of the links to another site.

Given that this issue is a red herring to some, I’d rather focus on what is new, unique, interesting, or controversial about this paper. …So what does the paper actually say?
Several previous (and perhaps flawed) studies have noted a correlation between globally-averaged low cloud cover and GCR intensity (and are cited in this paper). Whereas these papers have focused primarily on the cloud effects due to the 11-year solar cycle modulation of GCR intensity, this paper focuses on cloud effects due to a region of anomolously weak magnetic field (and correspondingly high GCR intensity) known as the Southern Hemisphere Magnetic Anomaly (SHMA), or South Atlantic Anomaly (SAA). A picture of the SAA, from Wikipedia, is shown below.
Vieira and da Silva show that the correlation coefficient between the magnetic field (a proxy for GCR intensity) and LW net flux anomalies decrease from the inner core of the SAA towards the outer edge. Their figure 5 (here, for JGR subscribers) is the most convincing in the paper, to me. Overall, this seems to be an interesting result, and may prove useful in understanding the GCR-cloud connection. I don’t feel particularly qualified to assess the importance of this paper, but do have a couple of questions and comments…
First, why did they only look at the South equatorial Pacific for their study? This region is subject to significant ENSO related variability, which must be separated from any variability caused by magnetic field anomaly. They only partially address their reasoning for choosing this area by saying

The patterns in the Atlantic are more complex and seem to be biased by meteorological systems over the South America.

So they threw out all of the Atlantic data based on these grounds. A better justification for this should be given in my mind. I’m willing to give them the benefit of the doubt on that, but still, what about the Pacific North of 0°?
Also, I’m curious about what the implications are for the potential solar-cycle modulation of cloud effects? I suspect that the change in GCR flux between the inner and outer core of the SAA is much greater than the solar cycle (I don’t have data to back this up, so please point me to the literature if I’m wrong!), and even in the data presented here the effects of things like El Niño dominate variations in cloud fluxes. So if one were to extrapolate the effects seen here to the types of GCR variations expected from the solar cycle, how large is this effect? Obviously this is a difficult question, and is beyond the scope of a GRL article. But it is one to keep in mind in assessing the implications of this paper…

Tags: climate · cosmic rays · satellite

13 responses so far ↓

  • 1 Hank Roberts // Aug 4, 2006 at 8:20 pm

    Here’s the rainfall pattern, there’s a peak that matches, but there are others.


  • 2 Hank Roberts // Aug 4, 2006 at 8:33 pm

    Wups, that was just the latest month. Again, I don’t know if clouds imply rain — didn’t find a cloud dataset. But looking for patterns, I wonder. That magnetic anomaly is also the place El Ninos begin, isn’t it? And downwind of a landmass, like the Atlantic rainfall concentration, which doesn’t have a magnetic anomaly apparent. Just poking at this, not pretending any competence to understand it.

    Here’s a year of rain; maybe you can find comparable cloud info, knowing better what to look for?

    Here’s the non-Java interface:
    There’s also a Java/Javascript interface (from which the ‘back’ button doesn’t work, apparently).

  • 3 Hank Roberts // Aug 4, 2006 at 8:36 pm

    One more — animated, 2003-2006. Same caveats. The rainfall concentration comes and goes; now I’ll shut up and hope someone can find similar cloud data, to get back on track.


  • 4 BKC // Aug 11, 2006 at 2:55 pm

    Vieira and da Silva show that the correlation coefficient between the magnetic field (a proxy for GCR intensity) and LW net flux anomalies decrease from the inner core of the SAA towards the outer edge.

    What does this mean/imply (in layman’s terms)? That cloudiness is more correlated with the magnetic field at the center of the anomaly?


  • 5 seand // Aug 11, 2006 at 3:19 pm

    This means that with higher magnetic fields (as you go towards the center of the SAA), a higher proportion of the variability in cloudiness (related to LW net flux anomalies) can be explained by the magnetic fields.

  • 6 BKC // Aug 11, 2006 at 4:04 pm

    Thanks for the response. That is interesting. Probably premature to draw many conlusions from the paper, but it does appear to provide strong evidence for GCR modulation of cloud formation. Certainly enough to pursue the answers to the questions you asked.

    The difference in heating/cooling between the inner and outer region of the SAA seemed very large. It seems like one (besides me) could derive a (crude) cloud to GCR sensitivity from this paper and answer some of the questions you ask.

  • 7 Ferdinand Engelbeen // Aug 27, 2006 at 3:00 pm

    As posted in the other discussion about clouds as feedback:

    clouds act as a positive feedback for solar radiation: the small change in TOA radiation is negatively correlated with low cloud cover, see Fig.1 of Kristjansson ea.. This study discards the GCR-cloud connection (which is worse for recent years), but induces a reverse connection between irradiation and low level clouds…

    I have read that the GCR-cloud connection will be investigated (within a broader base) in the CLOUD project of CERN, at last, the funding seems to be OK, but the first results will be available some 6 years from now…

    For the connection between clouds and irradiation, I have a burning question, which may be solved by knowledged persons here…

    It is known that IR waves are absorbed in the upper fraction of a mm of the ocean’s surface. Some expect that this is immediately reradiated, and no/less heat is dispersed in lower surface layers and/or transfered into evaporation energy. That means that SW/insolation variations have more effect on ocean temperatures than GHG induced LW changes.

    If the same is true for cloud droplets, this in part can explain the observed change in cloud cover under solar irradiation variability.

    What is known about LW absorption/reflection/transmission for different types of clouds?

  • 8 Eli Rabett // Sep 3, 2006 at 6:26 pm

    Two points. First, the magnetic field is not a proxy for the GCRs, but a controller of them. In this case you need to know the GCR flux at the top of the atmosphere as well as the magnetic field. OK, that’s picky.

    Second (I can’t get behind the subscription wall), previous GCR theories have had a problem matching the vertical and latitudinal distribution of cosmic ray tracks, e.g. the effect should be higher at the poles and higher up in the atmosphere. You also have to make sure to take all the know cofounding problems out, e.g. El Nino, see Paul Farrar in Climatic Change (10/2000)

  • 9 Hank Roberts // Sep 24, 2006 at 3:18 pm

    Is this the Tonga Trench? The GRACE satellites show a gravitational anomaly in this area, I’m just eyeballing the pictures though.
    Can anyone say the exact location being discussed?


  • 10 seand // Sep 24, 2006 at 8:14 pm

    Hank, I’m not sure what that is.

  • 11 William Connolley // Sep 25, 2006 at 3:18 am

    Hmmm, interesting paper, but makes large claims. And to continue your quote “In this study, we analyze just the patterns in the eastern Pacific, which are associated to the ENSO phenomena.” – this is asking for trouble, as Farrar showed, confusion with ENSO is a problem.

    This becomes clearer with the comment: “Observe in the middle and bottom panel that the ‘‘tongue’’ in the intensity of the magnetic field is quite coincident with a region of intense negative forcing in the eastern Pacific, which is surrounded by a region of intense positive forcing”. Yes indeed, but this *is* a circulation effect, not a cosmic ray effect. Which means that all of figure 3 (which attempts to correlate cloud effect with fluxes) may well be correct, but its not telling you anything about the GCR effect!

    Figure 4 then shows the obvious 1998 enso effect…

    Figure 5 looks good (in fact it looks implausibly good until you realise its an area-mean correlation) but again, this is simply the ENSO correlation that Farrar identified (or at least it might be; I don’t see that the authors have done anything to sort it out).

  • 12 Hank Roberts // Sep 25, 2006 at 5:19 pm

    I’m going to have to pay AGU their $20 to see those figures. Thanks, William.

    On the GRACE gravity anomalies, this will make it clearer what they’re finding — my question is whether the gravity anomaly correlates with magnetism, cosmic rays, clouds …


    The bottom image on that page gives the longitude to match up with the first image at the beginning of this thread.

  • 13 Bill F // Jan 15, 2007 at 2:09 pm

    As an update to this post, check out a few links:


    In response to Eli’s comment, one reason why GCRs may not act the same on clouds at higher elevations or at higher latitudes is that the nuclei for water droplet formation may not be the limiting factor in cloud formation there. I would surmise that in those locations, moisture content and/or temperature would be a more important limiting factor, which would prevent the GCR connection to clouds from working the same way.


    The relationship between GCRs intensity at the top of the atmosphere and at a given location on the planet is related to the orientation and strength of the magnetic field at that point. While the assumption typically made is that the field is more or less uniform in strength and parallel to the earth’s surface over most of the earth and that as you get closer to the poles, the field lines get closer and closer to perpendicular to the surface and the field weakens. What the Gubbins paper shows is that the field strength is not uniform and that the assumption that the field lines are parallel to the earth’s surface except at the poles is flawed. The influence of the SAA on GCR flux is so significant that some sensitive satellites are shut down as they pass through the sky over it so that they are not damaged by the change in flux. If there truly is a GCR-cloud relationship (which appears more likely with the Danish findings), then the SAA would be a good place to look for it to show up.

Leave a Comment