On a March evening in 1933, the Newport-Inglewood fault ruptured violently along the Huntington Beach coast. The quake brought down scores of buildings from Santa Ana to Compton, with Long Beach hit particularly hard.
The Long Beach quake, the deadliest in Southern California history, focused attention like never before on the seismic dangers the region faces.
But a new study suggests that the quake may have been caused by another factor: Deep drilling in an oil field in Huntington Beach.
The study, written by two leading U.S. Geological Survey scientists in Pasadena and to be published in the Bulletin of the Seismological Society of America on Tuesday, also suggests that three other earthquakes, including magnitude 5.0 earthquakes in 1920 in Inglewood and in 1929 in Whittier, may also be linked to oil drilling.
The two government scientists, Susan Hough and Morgan Page, wrote the report after a review of nearly forgotten state oil drilling records. They discovered that the epicenter of some of the Los Angeles Basin's largest earthquakes between 1900 and 1935 happened shortly after significant changes were made in oil production in nearby fields. During this era, the Los Angeles area was one of the world's leading oil producers.
"It was kind of more of a Wild West industry back a hundred years ago, and the technology wasn't as sophisticated," Hough said. "People would just pump oil, and in some cases the ground would subside — fairly dramatically." That possibly changed stresses on underground rock that could have pushed earthquake faults to rupture.
The report's finding does not mean that oil drilling is causing earthquakes in Southern California today.
The study only focused on earthquakes between 1900 and 1935. Different scientists have looked at earthquakes during more recent decades and have not found any reason to blame oil production for triggering earthquakes more recently in the L.A. Basin.
The reason could be that oil drilling practices in the basin have changed dramatically since the years when oil was first discovered in this region, and today's techniques may be safer and thus unlikely to trigger earthquakes as they might have done long ago.
Nowadays, water is carefully used to replace the pumped-out oil, which prevents land from sinking and helps extract more oil.
Most important, by keeping the pressure on the fault balanced, there would be less of a chance of disturbing the fault to rupture earlier than expected.
"It is … probable that changes to industry practices have largely mitigated the hazard," Page, the coauthor, said in an email.
Besides, "since the aftershocks of the 1933 Long Beach earthquake died out, the L.A. Basin has been relatively quiet seismically compared to the early 20th century," Page added.
The Long Beach earthquake killed about 120 people and caused major damage throughout the region. Its epicenter was in Huntington Beach, which became home to the largest known deposit of oil in California when it was first tapped in the 1920s.
The idea that human activity can trigger damaging earthquakes has become widely discussed amid the sudden increase in significant earthquakes in Oklahoma.
Scientists there have linked the dramatic rise in earthquakes in Oklahoma to the injection of wastewater underground, done after an oil production technique known as fracking.
By shooting this wastewater thousands of feet into the ground, it can set off earthquakes on faults that haven't moved in a long time. The injection pumps wastewater into areas where oil has not been extracted, so stress underground increases.
While the injection of wastewater has become controversial, the practice has not caused earthquakes everywhere.
Despite very large volumes of fluids being disposed of in North Dakota, that state has not had the human-induced earthquakes that Oklahoma has experienced.
"If the Los Angeles Basin were like Oklahoma today, we would know about it. We're obviously not inducing magnitude 5 earthquakes on a regular basis," Hough said.
Further understanding what is and isn't a problem would help make petroleum extraction operations safer. If there is something humans are doing that is causing problems, Hough said, "then it's a hazard that we can potentially manage."
The idea behind the study came up after Hough stumbled across old state reports on oil field operations that precisely identify where drilling happened. She found that there was notable drilling activity very close to the epicenter of the Long Beach earthquake that had begun just nine months before the temblor.
So Hough and Page identified five earthquakes in Southern California between 1900 and 1933 that were magnitude 5 and above. One was offshore west of Santa Monica, and there was no evidence it was linked with oil production.
But for the other four earthquakes — in 1920, 1929, 1930 and 1933 — the epicenter was no more than a few miles away from where there was notable oil drilling in the three to nine months before the earth shook.
In the case of the 1933 Long Beach earthquake, the scientists discovered that the seismic event occurred after an oil well that wasn't producing much petroleum was drilled much deeper. Suddenly, it was producing far more oil.
Something similar was observed before the 1929 earthquake. "And again, if you look at where the production was concentrated … it was essentially smack on top of where the earthquake was centered," Hough said.
One promising implication of the study is that, if true, the L.A. Basin is not as naturally seismically active as it's currently believed to be, if the earthquakes were caused by oil extraction processes that are no longer used.
In other words: "Maybe geologically, the L.A. Basin could be safer than we have thought. It's a possibility, at least," Hough said.
The bad news? It could mean that humans can cause more damaging earthquakes than some previously thought were possible.
Previously, some scientists speculated that human-induced earthquakes had a limit at close to magnitude 6, since that's what has been observed in Oklahoma. But if human activity induced the 6.4 Long Beach earthquake, then it raises the possibility that there may be no upper limit.
Caltech seismology professor Jean Paul Ampuero, who was not affiliated with Tuesday's study, called the report "a nice piece of seismological detective work."
Still, because the earthquakes cover a period where modern seismic sensors did not exist, it is inherently more speculative than a study of more modern earthquakes, Ampuero said. "It's probably the best you can do with the data that's available today from that period."
7 p.m.: This article was updated with a quote from Caltech seismology professor Jean Paul Ampuero, details of a study published in February that identified earthquakes in the San Joaquin Valley in 2005 as having a significant chance of being induced by an oil company's injection of wastewater into the ground, and historical context of Huntington Beach as a major producer of oil in Southern California.