Italian scientists say that the Sun has stopped directly causing temperature patterns on Earth. While the energy it blasts through space certainly still warms us, we no longer feel the effects of slight changes in its power. That’s what Antonello Pasini from the Institute of Atmospheric Pollution Research in Rome and his teammates say, after testing the various possible causes of the Earth’s temperature patterns.
“Our findings tell us that the causal link between solar radiation and global temperatures has weakened in recent decades,” Antonello told me. “This suggests that the Sun’s influence, which obviously exists and is strong, is probably overwhelmed by some other factors that at present are becoming more important in driving temperature changes. Further research is obviously needed on this topic, but my perception is that greenhouse gas emissions and other human influences are now strong enough to ‘obscure’ the Sun’s influence.”
The amount of energy the Sun produces varies, usually in regular 11-year cycles, and this has been an influence on the Earth’s temperature for thousands of years. But in research published in January, Antonello, along with Alessandro Attanasio and Umberto Triacca from the University of L’Aquila, found surprising evidence that this is no longer the case. They had tried out a test best known in economics, called Granger analysis and used to find out whether one set of events causes another, on climate science.
“We found that in recent decades there was a causal link – in the Granger sense – between greenhouse gases’ radiative forcing and the behaviour of global temperatures,” Antonello said. “The solar radiation did not show a significant link of this kind.” But Antonello, Alessandro and Umberto didn’t include weather patterns that bring natural variability to our climate, like El Niño and the Atlantic Multidecadal Oscillation (AMO), in that study. Now, by including them in a research paper published in the scientific journal Environmental Research Letters on Tuesday, they’ve been able to pin down when the Sun’s role faded away.
Though the maths is hard, the idea of Granger analysis is relatively simple. Researchers try and create a prediction of one measurement – in this case temperature. To do this they start with past values of temperature, and one other measurement – in this case from a natural variability pattern. They then create a second prediction, with both of the first two measurements and a third – in this case either solar radiation or radiative forcing from greenhouse gases. If the second temperature prediction is better than the first, by a margin scientists call statistically significant, then the third measurement used can be said to ‘Granger-cause’ the temperature change.
Using natural variability data showed the researchers that the Sun’s influence on temperature patterns has been waning for decades. “Our results show that greenhouse gases’ radiative forcing always Granger-causes temperature, at least since the 1940s,” Antonello said. “Solar radiation Granger-caused temperature in the 1940s and 1950s. But then, since the 1960s the significance of this causal link has weakened gradually, and finally disappeared completely after the 1970s.”
The comparative simplicity of this approach adds another line of evidence to findings produced by climate models, Antonello explained. “The Granger technique permits a statistical cause-and-effect analysis, in a field where other analyses are quite difficult due to the complexity of climatic models,” he said. “Results from global climate models show that natural forcings such as solar radiation and volcanic emissions alone can’t simulate recent global warming, correctly. This suggests a lower importance of the Sun in recent decades. At the same time, these models are often criticised by climate skeptics for their unavoidable incompleteness and uncertainties, so taking an alternative point of view is useful.”
Antonello Pasini, Umberto Triacca and Alessandro Attanasio (2012). Evidence of recent causal decoupling between solar radiation and global temperature Environmental Research Letters DOI: 10.1088/1748-9326/7/3/034020