The climate hockey stick, a popular visual metaphor for climate change, has received considerable attention. It depicts a slightly cooling trend in the Northern Hemisphere from 1000 A.D. until 1900 A.D. and then swings sharply upward in the last 80 years. It was created using tree ring data for the older timeframes but not recently - after 1960, tree ring data showed a cooling trend so it was replaced.

Clearly that was not correct but if tree rings don't detect the modern warming trend, they might also have 'missed' warming episodes in the past - we know that the climate is not cooler now, that is not the issue, but if tree ring proxies are unreliable, it casts doubt on the whole onus of media and IPCC accounts since 2001.

San Francisco State University researcher Alexander Stine may have an explanation. Changes in tree-ring density in the Arctic may be evidence of changes in light intensity during the trees' growth, which might account for this tree-ring "divergence problem." 

Tree rings consist of a low density ring, which forms early in the growing season, and a high density ring that forms late in the growing season. In colder parts of the world, the dense latewood rings tend to be denser during warm years. Temperature records inferred from Arctic tree rings do a good job of tracking temperature up until the 1960s, but subsequent Arctic tree-ring densities did not keep pace with increases in temperature, a discrepancy that is called the divergence problem.

Climate scientists have been aware of the divergence problem for some time, and it was mentioned in the emails from the Climatic Research Unit at the University of East Anglia, where it became the focus of attention during the 2009 "Climategate" controversy.  "Hide the decline" even became a video spoofing the brand new PhD who got the reins of an IPCC chapter in 1999 after coming up with the hockey stick.

This divergence is not a problem for understanding modern climate change in the Arctic, since 1980 we have thermometers that can be considered accurate - it's simply that in order to use tree rings as a proxy for the past, there has to be a clear understanding what's happening in modern times.

With his colleague Peter Huybers at Harvard University, Stine set out to understand why tree-ring density was declining in the Arctic. One possible explanation, the two thought, might be changes in light intensity that affected the trees' ability to grow. Starting in the 1960s, the amount of sunlight reaching the Earth's surface has declined. Scientists debate the cause of this "global dimming," with many scientists attributing it to pollution particles injected into the atmosphere by human activity that deflect incoming sunlight.

The researchers tested whether global dimming might be responsible for the decline in tree-ring density in the Arctic. This idea had been proposed in the past, but scientists had not been able to test the hypothesis. "It's very hard to distinguish a record that's controlled by temperature from one that's controlled by light, because light and temperature tend to vary together. Sunnier days are usually warmer," Stine added.

To get around this challenge, the researchers took advantage of the fact that there are regional variations in cloud cover and light availability throughout the Arctic, allowing them to compare trees that grew in the brightest and darkest areas but in comparable temperature ranges. They found that divergence was largest in the darkest parts of the Arctic, where changes in light should have the largest effect.

The researchers used changes in tree-ring density following volcanic eruptions to confirm the findings. Major volcanic events such as the 1991 eruption of Mount Pinatubo in the Philippines also have spewed tons of light-scattering sulfur dioxide particles into the atmosphere, decreasing the amount of sunlight reaching the surface.

Their analysis for seven different tree species suggests variations in light intensity caused by volcanic eruptions and global dimming both affect tree-ring density, and this impact is greatest in the darkest Arctic regions. In the brightest areas, the divergence problem essentially disappears, and tree-ring density is most closely linked to temperature instead.

Stine said the findings could have implications for geoengineering proposals that would pump more aerosol particles into the atmosphere as a way to block sunlight and potentially cool a warming planet. The tree-ring study suggests that Arctic trees might not grow as much -- and thus not soak up as much atmosphere-polluting carbon -- under such a plan.