There’s a new paper on arXiv about climate change.

Abstract: The sun’s role in the earth’s recent warming remains controversial even though there is a good deal of evidence to support the thesis that solar variations are a very significant factor in driving climate change both currently and in the past. This precis lays out the background and data needed to understand the basic scientific argument behind the contention that variations in solar output have a significant impact on current changes in climate. It also offers a simple, phenomenological approach for estimating the actual-as opposed to model dependent-magnitude of the sun’s influence on climate.

Let’s look at a few of the figures, since that’s where most of the interesting stuff happens. Figure 1 (below) is a variety of reconstructions of the total solar irradiance (TSI). Also plotted in grey (without an axis) is the sunspot numbers. He says “[n]ote the correlation between sunspot number and total solar irradiance.”

I don’t see the correlation between the grey and the black curve. Do you? Some of the shorter - < 11 year - variation may be correlated with the sunspot cycle, but the larger amplitude variations are not. He conveniently doesn’t not provide a numerical value for the correlation. There is no doubt it is small, and he is relying on the reader to “chi-by-eye”. Similarly, I do not see a correlation between the sunspot data (grey) and the other reconstructions, nor with the measurements on the longer than 11 year periods.

In any event, the figure does not show the latest measurements of TSI. From Greg’s TSI page, we find the following figure:

We see that we are currently in a period of low solar output: TSI is lower, and sunspot activity is lower. Clearly if the sun activity were the major cause of climate change, then we should be experiencing global cooling right now. Fear not, it’s not the actual TSI that causes the temperature to change on the Earth, according to Marsh. From figure 2, the temperature anomaly is “highly correlated” with the sunspot cycle length.

I’m unsure why there would be a correlation between temperature and the sunspot cycle length. But I’ve seen this graph used before, and have been unable to reproduce it. Fortunately, he cites Friis-Christensen and Lassen (1991), so I should be able to go back and use their methodology. It does look like there are only 23 data points on this particular figure, so the statistical significance could be in question. I’ll check that out, and report back.

Moving on to figure 5, we see a graph showing the change in cloud fraction, the change in cosmic rays, and the change in the solar irradiance.

You’ll also notice that the data only spans about 11 years, the same as the solar cycle - perhaps this is coincidence, perhaps not. You’ll also notice that it only uses data previous to mid-1994. If we look at the recent data, as well as previous data, we get the following graph:

Figure 1.GCR counts from Climax (red) and the aa-index(blue). The straight lines show the best linear-fit against time estimated through linear regression. The GCR measurements are shown in solid black line, from which a trend of -180 +/- 253 counts/decade is estimated, and this is associated with a p-value (the probability of this being different to the null-hypothesis: zero trend) of 0.477 (not statistically significant at the 5% level). The aa-index is represented by the blue line, and the corresponding trend of 1.5 +/- 0.4/decade is associated with a p-value of 0.0002 (highly statistically significant). A regession analysis points to a clear link between GCR and the aa-index, and the analysis of variance yields R2=0.1466 and the p-value=0. The yellow line shows the global mean temperature from CRU for comarison. [Data source: http://ulysses.uchicago.edu/NeutronMonitor/neutron_mon.html , http://www.cru.uea.ac.uk/cru/data/temperature/ and ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA ]. Image and caption courtesy RealClimate.org.

There is clearly no correlation between GCR and global temperature. There appeared to be a correlation between the years 1984-1994, which happens to correspond nicely with the data that Marsh presents. By not using the most current data, Marsh is clearly cherry-picking here.

References:
Friis-Christensen, E. and K. Lassen (1991), Length of the solar cycle: an indicator of solar activity closely associated with climate, Science 254: 698-700.

Marsh, G.E., Climate Change: The Sun’s Role, arXiv:0706.3621v1 [physics.ao-ph]. [PDF]

(No Ratings Yet)

 Loading ...

Related Posts:

Trackback URI |