Southern California suffered a number of big earthquakes in the early 1900s, a pattern that prompted experts to declare the state an earthquake hazard. But new work shows some of the biggest temblors might have been caused by oil and gas production, not nature. The finding could ultimately change scientists’ predictions for earthquakes in the Los Angeles Basin, and how well they understand man-made, or “induced,” earthquakes around the country.

It is challenging enough for scientists to determine whether a modern-day quake is natural or induced, and even more so for one that occurred a hundred years ago. The tools they now use to measure earthquakes were not as sophisticated back then, and historic records are limited. So researchers Susan Hough and Morgan Page at the U.S. Geological Survey relied on a combination of old scientific surveys, crude instrumental data and newspaper accounts to piece together details of quakes in the early 20th century. “It’s not as precise as having seismic data, but that doesn’t mean it’s hopeless,” Hough says.

From those documents, they determined the quakes’ magnitudes, the location of their epicenters and other attributes. They eventually narrowed their investigation to the handful of big, damaging earthquakes that hit the region between 1900 and 1935, during the Los Angeles oil boom. (The largest temblors were the best-documented ones.) Hough and Page restricted their study to this time period because records were spotty prior to 1900, and other researchers had already looked for induced earthquakes in Los Angeles post-1935. The pair also dug through old state reports to get statistics on the oil and gas companies’ production volume, along with locations and depths of their wells. They then analyzed the industry and earthquake data to see if they could establish a connection between the two.

Hough and Page found that oil and gas production in the Los Angeles Basin may have caused four out of the five major earthquakes in the region during that time. The largest—the 1933 Long Beach earthquake—was magnitude 6.4, killed 120 people and caused $50 million in damage (in 1933 dollars). Their study, published today in the Bulletin of the Seismological Society of America, explains that the majority of those big earthquakes occurred close to oil wells, often soon after production began. In every case oil and gas companies had drilled the wells more than a thousand meters down, which was unusually deep for that time period. All of these factors, Hough says, provide evidence for a link between the earthquakes and oil and gas activity. But it is important to note the study does not prove any direct causation. “What they showed is that the conditions are such that the earthquakes could well have been triggered by oil pumping activity,” explains David Jackson, professor emeritus of seismology at the University of California, Los Angeles, who was not involved with the work.

Hough also says people should not directly compare the historic earthquakes in California with what is happening in places like Oklahoma and Texas now. In the Midwest today scientists say man-made quakes are largely triggered by injecting wastewater from oil and gas production down into deep disposal wells. That was not done in the early 1900s in California, however, so Hough says the temblors were likely caused just by taking oil and gas out of the ground. The California and Oklahoma quakes, she says, correspond to totally different mechanisms. In 2015 Oklahoma had nearly a thousand earthquakes of magnitude 3 or greater, up from an average of two per year between 1978 and 2008.

Hough’s study, however, could still have big implications. If oil and gas production did set off most of Los Angeles’s major earthquakes during the early 20th century, then seismologists may need to recalculate southern California’s rate of natural earthquakes. “It could be that some of the earthquakes we thought were natural were actually related to human activity,” Jackson says. “And it’s possible that we could have fewer earthquakes in the future than we've had in the past.” Obviously, that would be good news—especially in a city of about four million people. Hough has already begun investigating historic quakes in other parts of California as well as temblors that hit Los Angeles after 1935, so estimates could change even more.

The study may even tell seismologists something valuable about man-made quakes in places like Oklahoma. Scientists don’t know yet whether or not there’s an upper size limit on human-caused earthquakes. In Oklahoma the highest-magnitude earthquake (which hit in September) was 5.8. “People have suggested there might be a magnitude cap around 6, because of the size that's been observed in Oklahoma,” Hough explains. But if oil and gas production in Los Angeles set off the 6.4 Long Beach earthquake in 1933, Hough says it might suggest no upper limit—that induced quakes could have magnitudes as large as natural temblors.

Other seismologists agree this is important to figure out. “This study brings in the idea that oil and gas production activities can generate large-magnitude earthquakes,” explains Richard Allen, director of the Seismological Laboratory at the University of California, Berkeley, who also was not involved in the work. “We need to start recognizing that there's a growing body of evidence that oil and gas production activities can generate large-magnitude, damaging earthquakes. And that's something we should all take very seriously.”