Most days, it’s easy to forget that coastal California sits at the boundary of two tectonic plates—the Pacific and North American—which are slowly sliding by each other, creating the San Andreas complex of faults. It’s easy to forget that one strand, the Hayward Fault, runs the whole length of the East Bay, cutting under Berkeley and Oakland, just a mile from my house, and that there is a one-in-three chance that it will produce a devastating earthquake before I’m a senior citizen.

But then there are days like January 4, when a magnitude 4.4 quake struck. It hit in the evening, a couple hours after my wife and I had put the kids to bed. It was strong enough to make us wonder, for a few seconds, if this was the big one.

After it passed, we resolved to get another flashlight. My wife ordered MREs from a prepper site. A few days later, she sent me a map from the U.S. Geological Survey showing the epicenter of the earthquake. It was two blocks from our house. I rode my bike over to the location. By the looks of it, the quake had struck on the backside of Alta Bates Summit Medical Center, the place where our kids had been born, and a place that we walk by nearly every day.

Of all the spots on the Earth reacting to the forces of tectonic motion, this had been the one to go. Right there.

Or so I thought.

* * *

It is strange to contemplate the nature of our existence on Earth, the sphere. Go to the highest mountain and you’re 29,000 feet up, five-and-a-half miles. The crust of our planet alone is many times that deep. The mantle is 1,800 miles thick. Everything we see as topography is inconsequential at the scale of the earth’s deep geology.

Yet, this is where life makes sense, up here on top of the thin crust. Geologists study the troubling world by engaging with humble rocks, and assembling them into narratives about time.

A portion of USGS map 2342 (USGS)

In human terms, I live in north Oakland, near the border with Berkeley. In geological terms, I live west of the Hayward Fault, on rocks washed down off the hills in recent times. On the maps, geologists call it “Qhaf”: “alluvial fan and fluvial deposits (Holocene).” The Holocene is the most recent geological era, which includes the present. Most of the flatlands of the East Bay are made of similar stuff, which lies over the top of what’s known as the Franciscan Complex, a heterogeneous collection of rock dating to the time of the dinosaurs tens of millions of years ago.

East of the Hayward Fault rise the hills, which are considered part of the Diablo Range, a complex set of low peaks that have resulted from the pressure of the tectonic plates coming together on the coast. Their geology is far more complicated, with different types of rocks bent and knuckled into place.

“Oakland is a real upstairs-downstairs city,” says Andrew Alden, a science writer, formerly of the USGS, who focuses on Oakland’s geology in particular. “And the fault is what separates it.”

Consider that for thousands of years, the areas on either side of the fault have been sliding past each other at a rate of a few millimeters a year. On geological timescales—like 100,000 years—a hill that would have once been adjacent to my house could now be half a mile south of it.

Troublingly, the recent observed “creep” of the plates past each other along the fault is less than half of the long-term average of about 10 millimeters. That difference will have to be made up for, eventually, and that’s what earthquakes do.

Faults are as narrow as the lines we use to paint them onto maps. They might run for tens or even hundreds of miles, and drop miles down into the crust, but they are not very wide at all. Alden said to think of faults almost like a bedsheet. Geologists have begun modeling this three-dimensionality, as in the following diagram of an area where the Hayward and Calaveras faults come together, near San Jose. (In fact, these diagrams suggest that the two faults should be considered as one.)

Hayward and Calaveras faults (Janet Watt)

The three-dimensionality of geology is essential for understanding earthquakes. To the average person on the surface of the Earth, the epicenter means everything. But that’s just the point we mark on our human maps by drawing a line from the hypocenter—the actual point in the Earth where an earthquake begins—up to the surface.

“Your whole premise is that the epicenter is important,” Alden said. “And it’s not. Here we are at the surface. The hypocenter was 12 kilometers down. You were right over it.”

So, really, I was 12 kilometers, or seven-and-a-half miles, from the actual spot the earthquake began. “I’m five kilometers away from you,” Alden says, forming a triangle with his hands to show that his distance from the quake would be the hypotenuse of the triangle formed by five surface kilometers and 12 kilometers of depth. “By the Pythagorean theorem, that’s 13 kilometers,” he said. “So, I was only a kilometer farther away from it than you were.”

The Earth is unfathomably huge. Us surface dwellers do not have great intuition for its scale or function. Human physical capacities—our size and the strength of our muscles—determine how we classify the world around us.

“We don’t have a lot of visceral insight into the Earth. Earthquakes remind us that rocks aren’t really solid,” Alden said. “Rocks are not solid. Nothing is solid. It is extremely hard rubber is what it is. It can flex. You can squeeze it. You can store energy in it. We’re not used to thinking of rocks like that.”

* * *

Now, about that epicenter. As I was scouring the Berkeley geological-sciences map library for references to my neighborhood, I got in touch with David Schwartz, a USGS geologist. He had just come out of an informal meeting about the earthquake with other researchers.

“While the location that is going out to the public and that is on Google Earth puts the epicenter very, very close to you, there are different ways to plot it,” he said. “The preferred location of the earthquake was really about 1.4 kilometers to the east, right on the fault itself.”

The earthquake had not been next door! He sent me a map of the situation. The green square is the public location, right near my house. The red square is the epicenter as calculated by another statistical method, called double difference, which locates it right up by the Claremont Hotel, right on the fault, and in the range of many other small earthquakes (the gray boxes) on the Hayward.

The public location of the epicenter is green. The epicenter, as calculated by the double-difference method, is in red. (Lind Gee)

Earthquake data seems very precise. There is a vast seismic-station network that records quake activity, all of which gets synthesized into the data the public can see. And the number-one thing people want to know is where the epicenter was. “In a sense, the epicenter has become sacrosanct,” Schwartz said. “Most people, when they look at how an earthquake is plotted, never really think about what’s going on at depth.”

Scientists have a more complex understanding of the data. “The location and magnitude of the earthquake usually get revised multiple times in the first few hours after an event—and can be revised even days or months (and even years in some cases) later,” said Lind Gee, the head of the Earthquake Monitoring Project of the USGS’s Earthquake Science Center in Menlo Park.

One reason is that what we see on the USGS website is automatically generated. Software sucks in data from the various monitoring stations and then spits out seismic events. “While the automatic systems are good, they are not perfect and we have a review process,” Gee told me. Once the humans look, the location and magnitude almost always change, even if only slightly.

In this case, the original, auto-plotted location was up by the fault. As more data from more systems came to be included, the automatic location got moved west, right by my house. But then, the double-difference method, which relies on many earthquake measurements with the same instruments over time, was applied, and the epicenter popped back over to the Claremont Hotel parking lot, right on the fault.

Even the magnitude of a quake is not so easy to pin down. For larger quakes, the USGS reports three different numbers: duration magnitude, local magnitude, and moment magnitude. “Even when we calculate multiple magnitudes,” Gee noted, “only one is ‘preferred’ and reported as the official magnitude.”

For me, all this data is an effective distraction from the reality that a great calamity will strike us eventually. I’m not alone in this. It’s as if we believe that if we can measure it, we can manage it. We can’t. The Earth is going to do what it does at depth, and our works on the surface may not survive.

“Down the road, when we do have the next large urban earthquake, people are going to be tremendously surprised at the damage that ensues,” Schwartz said. “Many of the buildings that were constructed were built before there were any real building codes, and in the case of the Hayward, just about every kind of infrastructure we have crosses the fault: BART, transmission lines, water, East Bay Municipal Utility District.”

In a real worst-case scenario, fires would break out while the water system was down, spurred by dry conditions and perhaps whipped by high winds. “There could be a tremendous conflagration. I don’t want to sit here and talk about doom and gloom, but we just haven’t seen, since 1906, sort of a worst-case scenario,” Schwartz said. “And we will.”