Skip to content

If one was asked to list the natural factors causing variations on the global temperature of the Earth over timescales from months to millennia I think most would write down such things as the oceans, atmosphere, sun, Earth’s orbit and their associated timescales. How many would, I wonder, include changes in the rotation of the Earth as an influence in changing global temperature. Certainly not the IPCC. In AR4, their 2007 review of the science of climate change, it doesn’t receive a mention. Yet it might be an influence, and an important one at that.

We have very accurate and precise measurements of the Earth’s rotation. Very long baseline interferometric techniques with radio telescopes observing celestial point sources can determine the length of day to 0.02 ms which is equivalent to less than 1 cm at the equator. Before radio data optical observations can be used, though at reduced accuracy.

The Earth’s length of day varies over a range of timescales and is thought to be due to interactions (angular momentum exchange) between the solid Earth and the atmosphere (the oceans have a much smaller effect) as well as, on longer timescales, interactions between the mantle and the core. However the observed decadal variation, approx. 4 ms, is too large to be explained by atmospheric interactions (typically 1 ms) and too swift to be due to anything inside the Earth.

Surprisingly, to some, there is a correlation between the length of day (LOD) and the Earth’s surface temperature.

The cause of this correlation is unknown. Some speculate that it is due to the Earth’s magnetic field. In particular rotational geomagnetic modulation of near-Earth charged particle fluxes that may influence cloud nucleation processes and hence planetary albedo on regional and global scales has been suggested. Such a mechanism is also being investigated by Henrik Svensmark of the Danish National Space Centre.

A recent paper by Dickey et al published in the Journal of Climate looks at changes in the LOD as well as the core angular momentum (CAM) and compares them to changes in surface temperature.


Their fig 1 summarises their findings. Click on image to enlarge. They use two temperature curves. One from Nasa GISS and the other a curved “corrected” for the influence of Anthropogenic Global Warming (AGW.) It is a fascinating graph. Note that the LOD and CAM are very similar, as models and the way it is calculated would expect. Note that between 1900 and 1930 the LOD and the temperature are correlated, although not for the 20 years before.

After 1930 something interesting happens – the temperature and the LOD part company. There is a 20-year spike in temperature that is not seen in the LOD data. Many have commentated on this feature. Some believe it is instrumental.

Post-1930-50 there is still a correlation between the observed temperature and LOD in the sense that they share gross features though not as exactly as before. However, when the AGW modified temperature is used the correlation is better and hold up to just beyond 2000 (note; it is fascinating to see the data for the 2000 – 2010 when the annual global average temperature hasn’t changed significantly. LOD data for this period shows it has been fairly constant, see fig 2. Click on image to enlarge.)


The authors say this break in correlation shows the beginning of AGW, which they put about 1930. This is several decades earlier than the onset of significant AGW as estimated by the IPCC that suggest it occurred in 1960 – 80, see fig 3. Click on image to enlarge.


1930 is a strange date for this disengagement. According to HadCrut3 the global temperature at that time was – 0.3 deg C below the 1961 -90 average, or about 0.7 deg cooler than the current period. Indeed, 1930 is not statistically significantly different from the temperature since the instrumental records began in about 1850.

If this is true then it has some interesting implications. It suggests that the excellent fit of pre-1960 data in fig 3 is spurious and that frequently made claims that climate models can reproduce accurately the Earth’s climate in the past decades are wrong.

If it is 1930 when the change occurred a look at HadCrut3 shows that for the first 50 years of AGW the temperature did not change significantly. This leads to the curious situation that in the past 80 years (taking into account the 2000 – 2010 temperature standstill) the number of years in which the temperature was higher than the previous year in the 80 years since AGW became a significant factor on the Earth’s temperature was 20 (or slightly fewer) or only 25 per cent of the time.

Another curious thing is that there is no indication in 1930 of any change in the behaviour of HadCrut3. There is a sustained, uninterrupted rise between 1910 and 1940 of about 0.12 deg per decade (similar to the 0.13 deg per decade if one looks at the recent post-1980 warm period, 0.2 deg per decade if you consider 1980 -2000.)

It is often said that prior to AGW taking hold solar forces was the primary driver in global climate change and that the transition occurred about 1950. This latest research, with its significant and intriguing correlation between LOD and global temperature at the time when solar influences were said to be dominant shows us once again that the science isn’t settled.