Apologies for a sign error in the previous version of the first paragraph.
The fluctuations of the cosmic ray flux may occur due to the variable galactic environment as well as the solar activity: a more active Sun protects us from a part of the cosmic rays. It means that a more active Sun decreases the amounts of low-lying clouds, which means that it warms the Earth.
Because the low-lying clouds remove 30 Watts per squared meter in average (over time and the Earth) or so, one has to be very careful not only about the very existence of the clouds but also about the variations of cloudiness by 5% or so which translates to a degree of temperature change.
A systematic effect on the clouds - e.g. one of the cosmic origin - is a nightmare for the champions of the silly CO2 toy model of climatology because the cloud variations easily beat any effect of CO2. Two alarmists, Sloan and Wolfendale, wanted to rule out Svensmark's theory by looking at the Forbush decreases, specific events of a solar origin named after Scott Forbush who studied them 6 decades ago, involving the plasma. However, their paper was incorrect.
In April 2008, this blog published the following relevant article:
In order to see the "tiger in the jungle", using Svensmark's words from a press release
Melbourne which already has August), and in order to separate these clean effects from the huge meteorological noise, one needs to increase the temporal resolution to several days and also cover the whole globe to dilute the effects of local weather.
Tomorrow, on August 1st, 2009, Geophysical Research Letters will publish a new paper by Henrik Svensmark, Torsten Bondo, and Jacob Svensmark:
The People's Voice (summary of the paper)When you click the second link above and obtain an error message, press alt/d and enter to reload the URL: without a direct external link, the PDF file will be displayed correctly. Or open the Google cache as PDF-like HTML.
Cosmic ray decreases affect atmospheric aerosols and clouds (full text).
Svensmark and his collaborators have looked at 26 Forbush events since 1987 (those that were strong according to their impact on the spectrum seen in the low troposphere where it matters): most of them occur close to the solar maxima (in the middle of the 11-year cycles). The observations with a much better temporal resolution imply that the mass of water stored in clouds decreases by 4-7%, with the minimum reached after a nearly 1-week delay needed for the cloud nuclei to get mature. Roughly three billions of tons of water droplets suddenly disappear from the atmosphere (they remain there as vapor, which is more likely to warm the air than to cool it down).
An independent set of measurements has also shown that the amount of aerosols, i.e. potential nuclei of the new clouds, also decreases. All these "strength vs decrease" graphs display a lot of noise but the negative slopes are almost always significant at the 95% level (with one dataset being an exception, at 92%, which is still higher than the official IPCC confidence level that climate change is mostly man-made).
Each Forbush decrease can therefore warm up the Earth by the same temperature change as the effect of all carbon dioxide emitted by the mankind since the beginning of the Industrial Revolution. While you might think that such an effect is temporary and lasts a few weeks only, it is important to notice that similar variations in the solar activity, the solar magnetic field, and the galactic cosmic rays take place at many different conceivable frequencies, so there are almost certainly many effects whose impact on the temperature - through the clouds - is at least equal to the whole effect of man-made carbon dioxide.
The supplementary material attached to the paper includes many rather sophisticated integrals, power laws, and detailed graphs and numerical checks.