Prof. Qing-Bin Lu published a paper (DOI: 10.1103/PhysRevLett.102.118501) in Physical Review Letters suggesting a correlation between Cosmic Rays and Ozone Depletion, using data up to 2007. There are predictions on the paper of large Ozone losses in 2008-2009, and certainly Cosmic Rays have peaked in 2009 as a result of the prolonged minimum in solar activity. (Solar activity tends to shield the earth from cosmic rays).

This article has a demonstration of the depletion and recovery of Ozone, which does seem to correlate with F10.7 solar flux.

Lu’s paper says little about climate. Ozone seems to have an indeterminate effect as a greenhouse gas. What is clear is that Ozone reached a minimum around 1997, and has been recovering slowly since. The solar minimum could be expected to slow that recovery. As it stands, Ozone seems to be inversely correlated with surface temperature, which has a recent peak around 2000, and a more recent stabilisation. Actually, Ozone may well be driven by temperature so the whole argument seems a little confused.

Now CFCs are strong greenhouse gases, so at least that part makes sense. CFC production started roughly at the point where temperatures started to rise, and cosmic rays could impact the lifetime and concentration. Nothing proven yet, just more potential mechanisms.

A 2007 presentation (pdf) has more detail on modelling the recovery of ozone, and the presence of CFCs.

I postulate that in addition to the peaks in F10.7 associated with sunspots, there is also a more steady increase from features which are not directly observable. Sunspots are more strongly quantised, and even taking the monthly averages, there is a noisy signal which is difficult to interpret. I originally plotted this graph to demonstrate a couple of months afo that the sun was dragging itself out of the current minimum, and also to show how a delay in the onset of SC24 would affect the shape of the curve. The graph shows the equivalent SC23 ramp-up (adjusted to fit the decline of 22 with the decline of 23).

Note:

1. The signal is noisy, it is difficult to estimate trends
2. SC24 is making headway, but maybe half as fast as 23.
3. SC24 is 2 years delayed in getting to this amplitude compared with 23.

Click for full-sized image.

Note that I am not making any claims that this is having an effect on climate. Even another 5 years of data alone will probably be insufficient to give that claim any credence.

In order to construct this chart, I took the approach used when measuring the noise figure of sensitive radio amplifiers. I assume that the measured value is the sum of two input sources, the background level, which here i assume to be 67.0 (this happens to be the lowest monthly observation in my time-series).
I then construct an inferred cycle activity value which is the amount i have to add to the background in order to reach the observed value. Since we are adding powers of un-correlated signals (really two noise sources) the summation is done a the sum of squares. (The reasoning being that if we convert to voltages, voltages add linearly).
I took an arbitrary decision to draw the two sides of the minimum as being 100% contributed by the fading cycle 23 and the emerging cycle 24, but hopefully it is clear that if I wanted to decompose the 3.6 flux units in Nov08 into cycle 23 and 24 evenly (each giving 2.6 units) and then subtracted the same cycle 24 value from Oct08, I would get the following split for Oct08

Now, the period of increasing flux is short, but the rate of increase is not wildly different to the decrease. What I have shown here is a better method of analysing the small variation in the measurement, rather than making any attempt to explain the underlying physics or make predictions. Just because the curve appears to be almost flat and quite noisy doesn’t mean that there is no information to be gained from it.

Update 13July

In order to confirm that the ramp is not very differnet from the start of Cycle 23, I have overlaid the same decomposition of old and new cycle for the previous minimum. I took an arbitary start point for the alignment, and plotted the end of cycle 22, and the start of cycle 23. in order to aportion flux between the two cycles, at an arbitary point I replaced the cycle 22 values with a simple straight line (as indicated in the dotted section) and subtracted this from the total, leaving a nominal strength for cycle 23. Both of the tail ends of old cycles show a consistent form, it is too early to tell how the new cycle compares with the start of cycle 23, but it is looking good so far.