The King Fire in Northern California, taken by NASA’s Aqua satellite on Sept. 17, 2014. The fire, caused by arson, was started on a Saturday. (AP Photo/NASA)

This story has been updated.

surprising new scientific paper suggests that the pattern of global wildfires varies based on the day of the week — with considerably fewer fires globally on Sunday than on other days. And not only that: It partly attributes this pattern not to anything natural about ecosystems, but rather to human behavioral patterns — including weekly rituals that are ultimately rooted in culture and faith.

“Our weekly routines are based on religion originally,” says Nick Earl of the University of Melbourne in Australia, who published the research in Geophysical Research Letters with two colleagues. “And you can see that in the weekly cycle of fires.”

To show as much, the researchers used NASA satellite imagery to look at all global fires from 2001 to 2013, with images taken four times each day. The images were then scanned by an algorithm to detect large fires, which can occur both for natural reasons — i.e., lightning — but also due to human causes, including arson and accidents, deliberate forest management, and the lighting of fires for forest clearing and agriculture (the apparent source of the Indonesian peat fire catastrophe that occurred this year).

This appears to have been the first time that researchers tried to detect a weekly cycle in the occurrence of global fires — which, if fires were a purely natural phenomenon, shouldn’t exist. “There’s nothing in nature that’s on a weekly cycle,” Earl says.

Yet after applying various statistical tests to the very large set of data, the researchers found that while patterns varied by region, the results on a global scale showed a considerable drop in fires on Sunday, as well as a weekly peak on Tuesday. “There were a total of 104 million Sunday fires globally over 2001–2013, this being 9 million (8%) fewer than the Tuesday total,” the researchers noted. This pattern was particularly pronounced in some parts of the world, including the United States, and less so in others.

“Globally, there’s a very clear weekly cycle of fires, Sunday being the minimum,” Earl says. As he notes, the data include all the naturally caused fires too, from sources like lightning strikes, which would tend to suppress any human signal.

Here’s an image Earl provided, visualizing some of the results (click to enlarge):

Credit: Nick Earl.

Of course, it’s one thing to find such a pattern, and another to interpret it. Still it’s hard to ignore that Sunday isn’t just any day — it is the day of rest in many societies around the world.

Sure enough, the paper makes the attribution to religion directly: “Our results show that weekly cycles in active fires are highly pronounced for many parts of the world, and these cycles are strongly influenced by the working week and particularly the day(s) of rest linked to religion,” the authors write.

Asked if he is confident in this attribution, Earl simply said that he’s trying to explain the data and doesn’t see another way of doing so. “If someone comes up with a different idea, that’s great, because I can’t think of one,” he said. “I can’t think of any other reason for it.”

Guido van der Werf, an expert on global fires at VU University Amsterdam, praised the study in a comment to the Post, noting that “It confirms the important role of humans in starting fires in the tropics.”

“Keep in mind that an important role for humans in starting fires does not mean that without these human ignitions there would be no fires, new research indicates the opposite may be true,” said van der Werf.

The research also hinted that, while fires on a global scale go down on Sunday, there might be regional variations, also tied to faith — including different practices in non-Christian countries. For instance, Earl points to predominantly Muslim Kazakhstan, where the fire minimum was actually on Thursday and Friday.

“Areas with a higher Muslim population like the Kazakhstan region have minima on Thursday and Friday,” a fact sheet provided by Earl notes. “This is likely to be because Friday is ‘the day of assembly’ and prayer for the Muslim faith, so there is less industrial activity on these days.”

However, it’s not clear how far you can take this: The paper notes that the data for Kazakhstan were not statistically significant, although it also suggested some reasons “any significant signal may be being masked.” Another predominantly Muslim country that sees many fires, Indonesia, also failed to show a statistically significant late week minimum. When it comes to such countries, then, the data provide just a “hint that it goes against sort of the global pattern,” Earl says.

All of which suggests both that there are other factors involved across countries that determine patterns of fires, and that there may be limits to the technique and the inferences that can be made from it. The study’s strongest finding, then, is simply the global picture, as well as the presence of weekly cycles in a number of regions.

If the research is correct, it wouldn’t just be a surprising observation — it would have consequences. For even as humans shape weekly fire cycles, says Earl, those weekly fire cycles feed back and shape the weather we all experience. Fires, after all, emit large volumes of aerosol particles, which can have complex atmospheric effects that can affect both temperature and rainfall.

The most striking result, though, remains the attribution of fire patterns, at least in part, to culture and religion. Asked if this means we can see human religion from space, Earl put it like this: “I never actually thought of it like that before. You kind of are.”