Photo: G.K. Gilbert (U.S. Geological Survey Photographic Library)

Researchers used a grainy photograph of a toppled train combined with an eyewitness account to analyze the deadly earthquake that struck San Francisco over 110 years ago.

In 1906, a magnitude 7.9 earthquake struck San Francisco, destroying much of the city and killing as many as 3,000 people, making it one of the deadliest earthquakes in history. But it also gave birth to modern American earthquake science. Now, researchers are trying to understand the quake using cutting-edge techniques—and some old documents.


“It’s important to understand how these larger quakes occurred so we can build safer buildings,” explained Swetha Veeraraghavan, postdoctoral research associate in structural computational science at Idaho National Laboratory. “I think these types of eyewitness accounts really help with this. That’s why I found the problem really cool.”

Scientists have tried analyzing old images and eyewitness data before in order to better understand the 1906 quake. Newer earthquake simulations based on more data and better computational power inspired the researchers to take another look at this famous image of a steam locomotive sitting on its side, taken shortly after the quake. They also examined a conductor’s description of the incident. The 1907 account reads:

At Point Reyes Station at the head of Tomales Bay the 5:15 train for San Francisco was just ready. The conductor had just swung himself on when the train gave a great lurch to the east, followed by another to the west, which threw the whole train on its side. The astonished conductor dropped off as it went over, and at the sight of the falling chimneys and breaking windows of the station, he understood that it was the Temblor. The fireman turned to jump from the engine to the west when the return shock came. He then leaped to the east and borrowing a Kodak he took a picture of the train here presented.


This might sound like meager information with which to reconstruct an earthquake, but it provides more data than you might think. The quake must have been powerful enough to cause the rocking and the tipping of the train. The team simplified the train’s measurements into a long rectangular prism, and simplified the quake to an up-and-down wave motion of the ground. They then used equations devised to describe a rocking rectangular block in response to ground motion to calculate how big the wave must have been in order to tip the train.

Their analysis revealed that the minimum acceleration of the ground required to overturn the train would have been 4 meters per second squared (a little less than half the acceleration of something falling to the Earth in a vacuum), while the minimum velocity would have been 0.5 meters per second, or around 1.6 feet per second.

On top of their analysis of the train tipping, the researchers were able to estimate the location of the earthquake’s hypocenter—the area where it actually occurred in the Earth below the epicenter on the surface—based on the train’s location and behavior. While previous models have suggested three locations for a hypocenter, this result reveals that only a hypocenter to the south of Point Reyes, California, would recreate the eastward, then westward tipping of the train, according to the paper published in Seismological Research Letters. This leaves a hypocenter offshore from San Francisco or near the California city of San Juan Bautista, confirming previous estimates.


The paper comes with uncertainties—for example, the researchers didn’t know how much the train weighed, and are using a simplified model to understand what went on.

Still, it confirms a body of research recreating this important quake. And it shows more generally that even in the absence of data, context clues, eyewitness accounts, and photographs can prove important for understanding earthquakes. Veeraraghavan said: “People have even used tombstones or statues that toppled over to estimate the ground motion.”