Left without official measurements, scientists can estimate atmospheric particle data using colors in famous paintings.
Chichester Canal by JMW Turner. Image: Wikimedia Commons
Scientists are always trying to compare today's climate to historical ones in an attempt to find out what caused massive biological and cultural shifts and as a way to predict what might happen if our climate keeps changing. One of the major problems, however, is that there are very few tools they can use—we didn't have advanced scientific instruments monitoring the atmosphere back in the 1600s, after all. But it turns out that some clues are coming from an unexpected place: The landscape paintings of master artists.
It sounds outlandish, but researchers at Greece’s Academy of Athens have found that the colors of sunsets of famous paintings can be used to roughly estimate aerosol optical depth, a measure of particles such as dust, sea salts, volcanic ash, and smoke in the atmosphere. To be fair, aerosol optical depth is one of the more basic measures of climate and can be seen by the naked eye—more particles in the air will change the color of the sky. This is some pretty wild stuff, but take a look at these two photos, painted as a companion to this latest study:
Image: Atmospheric Chemistry and Physics
Christos Zerefos, lead author of the study, published in Atmospheric Chemistry and Physics, enlisted the help of Greek landscape painter Panayiotis Tetsis to paint two sunsets off the coast of Greece’s Hydra island both immediately before and immediately following what’s known as a Saharan dust event, which kicks particles in the atmosphere and is known to change a region’s AOD. But you don't have to be a trained painter or a trained atmospheric scientist to see that the color of the sunset is different in those two photos.
As might be expected, there was much less human pollution in the time he studied, between 1500 and 2000 (with a heavy bias towards the early 19th century), but there were still plenty of particles in the air, mainly as the result of several major volcanic eruptions. Zerefos found that in volcanic years—the year of an eruption and three years after—sunsets were notably different than those in “non volcanic” years.
Using several hundred paintings from London’s Tate Gallery and London’s National Gallery, Zerefos says that estimated aerosol optical depth derived from the red-to-green ratios in painting sunsets correlate with computer models and existing historical data, such as ice core information. In fact, the style of painting and the painter of any given work didn’t even seem to matter, especially when many paintings were averaged together.
“AODs from a multi-hundred sample of paintings show statistically significant correlations with independent proxies. Structural differences in paintings do not seem to alter the results,” he wrote in the study. “When averaged in 50 year intervals, AODs from paintings in non-volcanic years agree with completely independent data sets with the observed increases of the industrial aerosol in the past 150 years.”
It’s not the first time that art has been used to help out science in a historical context: A study published earlier this month suggested that ancient Greek plays could be used to infer something about the “halcyon days” of clear, sunny weather during ancient winters. It’s also not the first time Zerefos has tried to use paintings to study the atmosphere. His latest study confirms preliminary findings from a study he published in 2007. And it's not even the first time paintings have been tied to climate. A 2012 paper by researchers at German Research Centre for Geosciences suggests that art analysis has always been useful for scientists:
For as long as human beings have used pictures to express themselves and communicate, they have made their natural environment the subject of their depictions. Information about the weather and climate can be found almost
everywhere, with known early examples being the cave paintings at Lascaux in France and the “Swimmers in the Desert” at Gilf Kebir in south-eastern Sahara. The latter provides clear evidence that this part of the desert was once humid and green ...Thus, there are numerous representations of climatic phenomena in pictures, which directly raises the question of whether these representations would be suitable as proxy data for climate research.
There are some obvious questions that can be raised about all of this: What about color degradation over the course of a couple hundred years? What about different coatings and storage techniques? Why would we even want to do this? Zerefos writes that, because he's looking simply at the red-to-green ratio versus the absolute color of a painting, issues of degradation can be overcome, as can issues of painting coatings and storage. As for the "why," aerosol optical depth may be a basic measure, but it's a pretty useful one: Higher levels of aerosols are associated with cooler temperatures, because the particles reflect sunlight back into space. It's one of the reasons why aerosol seeding has been floated as a way of reducing global warming. AOD can be used in many climate models to estimate the general climate situation.
Though his historical estimates line up with what we had already guessed, he decided to do the modern test with Tetsis to see how accurate he could be. He found that estimates based on Tetsis' paintings closely aligned with actually recorded AOD measurements.
“During the two day experiment the substantial difference between the aerosol condition of the first and the second day provided a more adequate data set, supporting the assumption that a painter is able to reproduce such an aerosol change,” he wrote. “Quantitatively it has been proven that this assumption was correct.”
Of course, estimating aerosol content of the atmosphere in the 1600s is never going to be an exact science, and AOD is a relatively basic measure of what was going on at the time. But scientists will take what they can get, and it turns out landscape paintings are more than just pretty pictures.