CLOUD experiments

The first results from the CLOUD experiments have been getting a lot of media attention. The focus of the attention is the Nature paper that was published this week: [news][paper].

The goal of this project is to determine whether cosmic rays have a significant impact on clouds.

Let's boil this down a little. This project is a laboratory experiment at CERN. It is a cloud chamber, basically an isolated volume of air that is precisely controlled for temperature and pressure. They put very pure air into the chamber, add a little background water, and some gases like ozone, sulfuric acid, and ammonia. The chamber is heavily instrumented to look for nucleation, which just means that they try to keep track of particle formation that occurs as the vapors interact and possibly start condensing. They can do this in neutral conditions (like a classical cloud chamber), or they can shine a pion beam into the chamber. That beam is adjusted to mimic cosmic ray bombardment. The goal is to see if cosmic rays produce ions that enhance the formation of particles, which could then go on to become the seeds for cloud droplets.

The answer seems to be that shining that beam into the chamber does produce more particles. This actually isn't a surprise, as far as I can tell. One important point is that nucleation rates, that is the rate of particle formation, are smaller than observed rates unless the temperature is quite low. This means that it is unlikely that cosmic rays ionizing gases near the Earth's surface is a major source of particle formation. Certainly there is particle formation, but it is likely to be a small source of the total number of particles. This result may change when they start adding in organic molecules, but that is future work.

There is better coverage on RealClimate: link.

There is hubbub about this result because there is a crack-pot theory that galactic cosmic rays are a major control of climate because of their impact on cloud formation. There are major flaws with this theory. My own take is that cosmic rays probably do produce some of the particles in the atmosphere that go on to become cloud condensation nuclei, but there are many paths to becoming cloud condensation nuclei, and there are lots and lots of these particles around. In fact, I seriously doubt that cloud formation is frequently affected by the limitation of these aerosol particles. I've been thinking about this in terms of observed cloud properties. The number of cloud droplets is connected to the number of aerosol particles available: over land where there are lots more aerosol particles, there tends to be more, smaller droplets in clouds, while over the remote ocean the clouds are made of fewer, larger droplets. In very polluted conditions, we can observe changes in the cloud properties that follow that same trend. I think the downfall of the cosmic ray theory of cloud formation comes from the fact that out in the middle of the ocean there are still tons of aerosol particles. While many of those particles may come from cosmic ray influenced nucleated vapors, there is no evidence that there is a shortage of other sources of aerosol, so if the intensity of cosmic ray bombardment were to change, it seems unlikely that other sources of aerosol wouldn't fill whatever tiny void that change would make.

Besides this basic criticism (which amounts to the originators of the theory simply having a bit of a myopic view of cloud formation), there is also a clear lack of evidence for cosmic ray intensity modulating cloud/climate. The RealClimate piece covers that. Finally, there is the link to climate change, for which there is absolutely no evidence.

So my summary would be something like: This research presents experimental results that suggest that ionization by cosmic ray-like effects can impact nucleation rates in conditions similar to the Earth's atmosphere. The role of such nucleation enhancement in the Earth's atmosphere remains unclear, especially given that the impact seems most pronounced in conditions that are outside the atmospheric boundary layer. This is a nice contribution to basic aerosol research, which should help to constrain models of aerosol formation. The impacts on cloud formation and the Earth's climate can not be assessed with the data collected so far.


The authors are only slightly overselling their results, which is typical for authors of Nature papers. The lead author's comments can be heard in the embedded YouTube clip. The media coverage, and especially the climate change denier blogosphere, is lighting up like this experiment proves something controversial. It does not.

Reactions to the Spencer paper

It's a very bad paper. That is the short story. I haven't thought through all my criticisms of it yet (maybe I'll write something here eventually). In the meantime, I like this overview piece at Climate Central by Michael D. Lemonick [LINK]. I especially enjoyed the comment that, "... it's not that NASA data are blowing a hole in anything. It's that Spencer's interpretation of NASA data are blowing... something, somewhere."

For those of you looking to actually read the paper, it is in a journal called Remote Sensing, and it is open access. You can find it by looking up doi:10.3390/rs3081603. Let me reiterate that this is a bad paper, with many incorrect statements, assumptions, and reasoning. It isn't worth you time reading this paper when you could better spend it reading an informative one about climate sensitivity... oh, I don't know, maybe doi:10.1175/2008JCLI1995.1