Scientists are using "big data" to study the decline of stromatolites. Big data could help with synthesis and research on many subjects, including climate change.

Stromatolites in Shark Bay, a World Heritage Area in Western Australia. Photo: Shutterstock

Stromatolites are naturally occurring structures in the world’s oceans, built up by lime-secreting cynobacteria. If you haven’t heard of them, it might be because they’ve been pretty uncommon since the Precambrian era, just before the rise of multicellular life. They do still form in shallow lagoons, but much less often than they used to.

Paleontologists used to believe that the decline in stromatolites was due to the fact that they were microbes, and they were being eaten by larger creatures. But a new study has found that correlation to be pretty weak, while a stronger one is between the decline of stromatolites and the change in seawater chemistry, as the oceans contained less carbonate.

“The best predictor of stromatolite prevalence, both before and after the evolution of animals, is the abundance of dolomite in shallow marine sediments,” said Jon Husson, a post-doctoral researcher and co-author of the study.

Dolomite is a high-magnesium variety of carbonate. Carbonate is the type of sediment that forms limestone. According to the researchers, dolomite is harder to make than low-magnesium carbonate, and it’s very rare—it’s seen in only a narrow range of marine environments.

The study was possible because of machine reading, using a system called GeoDeepDive. The program was able to find references to stromatolites in 10,200 papers. Of those 10,200 papers, the program designed by researcher and study third author Julia Wilcots was able to extract 1,013 that also included the name of a rock unit. This allowed the researchers to link stromatolite occurrences to Macrostrat, a database describing the known geological properties of North America’s upper crust, at different times and depths.

Wide-ranging studies like this would have been all but impossible previously, but with continued improvements in databases and machine reading, scientists can do studies like this on a whole new scale.

“Now that this study has been done, we can run the stromatolite application again and again. We can refine the searches, and they will evaluate the new data that is being published all the time,” said Shanan Peters, a professor of geoscience at the University of Wisconsin-Madison and study first author. “So a rerun could make a better study, with minimal effort.”

Approximately 10,000 new published papers are being added to GeoDeepDive every single day.

The new study “allows us to do the kind of analyses that scientists used to only dream about,” Peters said. “If we could just compile all the published information on…anything!”

You can expect to see “big data” technology used more and more in studies of all types. New papers are being published every day about the causes and effects of climate change, but those studies are rarely capable to looking at the big picture of all of the combined effects of climate change. As scientists continue to discover new effects, they continually revise their understanding of how climate change is affected the world today, and their predictions of how it will do so in the future. All of that requires data, preferably from as many sources as possible, and that is a lot easier to manage with the help of computers—they can read far faster than any human ever could.