At a downtown coffee shop, the mugs begin to chatter.

Small talk stops. There’s an uneasy hush. A barista’s hand hovers over the bean grinder. Customers lift their eyes from phone screens.

Then the world rattles up and down. Dishes jitter off tables, shattering on the floor. Someone screams. Seconds later, it’s as if Whidbey Island is trapped in a cocktail shaker, lurching back and forth.

Brick chimneys cascade off rooftops. Bookcases and china cabinets topple, trapping people beneath. Disoriented drivers wonder what’s wrong with their cars, then realize something much bigger is amiss.

Facades crumble off buildings along Oak Harbor’s Pioneer Way, and some of the oldest structures in Langley and Coupeville collapse in a roaring cloud of dust. People stagger into the streets to avoid an avalanche of debris. Standing becomes almost impossible as the jolts turn to rolling waves.

Around Puget Sound, it seems everyone knows about “The Big One,” the potential magnitude 9.0 Cascadia Subduction Zone megaquake some scientists say is due any day.

But this isn’t it.

This earthquake is along the southern Whidbey Island fault, a less-known, less-studied subterranean boundary. And experts fear it could be even worse than “The Big One.”

The fault zone, known to geologists as SWIF, cuts through Puget Sound in a diagonal line roughly from Port Townsend to the southern tip of Whidbey Island, then to Mukilteo, Bothell, North Bend and possibly farther east below the Cascades.

Consider a magnitude 7.4 quake with Whidbey Island at or near the epicenter.

Under a scenario played out in a 2019 U.S. Department of Homeland Security study, state-maintained bridges would be severely damaged, leaving them unusable for months or years.

Buildings could sustain extensive damage. Power could be out for days. Restoring tap water to some homes could take over a year. Some residents may lose housing temporarily or permanently.

Hundreds could die, with thousands more injured.

This is a hypothetical scenario created by Mark Murphy of the Snohomish County Department of Emergency Management. In 2017, he began studying the possible aftermath of a major SWIF quake. He combed through state and federal data to understand the risks, and to help train first responders.

The threat to Puget Sound from a quake along the Cascadia Subduction Zone, off the coast of Washington, Oregon and California, is well documented. Experts believe a magnitude 9.0 could happen there anytime in the next 200 years or so.

Despite its location well offshore, a Cascadia quake would likely kill at least 10,000 and injure more than 30,000 in Washington, Murphy found.

A major southern Whidbey quake, on the other hand, could kill and injure many more, according to Murphy. Unlike The Big One, scientists who have studied the southern Whidbey fault have far less understanding of when the next sudden shift might hit.

In fact, until the 1980s, no one knew SWIF existed. Until much more recently, no one really understood what it could do to a region of over 4 million people.

A cryptic fault

A pair of scientists with the U.S. Geological Survey first theorized that a fissure between two major blocks of the earth’s crust might run through this slice of Puget Sound. Years ago, Howard Gower and James Yount came to the Puget lowlands to study earthquake risks and stumbled on what appeared to be a fault in Island and Snohomish counties.

In 1985, with little concrete evidence of its existence, the pair included the possible fault on a geologic map published by the USGS.

They didn’t recognize the significance of what they found.

“They knew something was there,” said Sam Johnson, a retired USGS geologist who would follow up on their work. “But they didn’t document it hardly at all.”

So the fault remained mostly a mystery until the 1990s. Johnson, on a whim, acquired the data that would prove its existence beyond a doubt.

At the time, Johnson worked in southwest Washington, searching for natural gas and oil deposits. In the late 1960s, speculators considered the Puget Sound region a frontier for petroleum exploration. Oil companies descended in search of riches. Finding nothing of serious monetary value, the companies abandoned reams of information they had gathered through seismic surveys.

Years later, that data proved priceless.

“It might as well have been sitting in a drawer,” Johnson said.

Though it was not directly related to Johnson’s work, he asked a friend working for Mobil Oil to pass along the information. The friend obliged.

“It’s just the way scientists work,” he said. “If they know there’s data available that could help them in any way, they want to get it. It’s a natural curiosity.”

Johnson’s curiosity changed the course of his career.

The seismic mapping had cost millions of dollars — far beyond what most geologists on a government budget could scrape together.

Like a sonogram, the seismic surveys allowed Johnson to see outlines of massive fissures in the earth’s crust. Within minutes, he spotted something groundbreaking.

“Once we got it, we were sort of shocked to see these big faults in the Puget lowlands,” he said.

The southern Whidbey Island fault, and several others, were exposed for the first time from a camouflage of forest, ocean and glacial sediment. It startled Johnson that such massive faults had gone undetected for so long.

Not all faults are active. But the mapping offered geological clues that the newly found fault was indeed capable of future quakes.

Geoscientist Brian Sherrod stands near the submerged southern Whidbey Island fault line at the Brightwater Treatment Plant in Woodinville. (Andy Bronson / The Herald)

In the early 2000s, USGS scientists including Brian Sherrod set out to further Johnson’s work and better understand the slumbering fissure. The key, Sherrod’s group would discover, was buried on Whidbey Island under layers of mud, peat moss and decaying marsh grass in the murky tidal waters at Crockett Lake, alongside the Coupeville ferry dock.

Sherrod remembers his son, age 5 at the time, playing with toy trucks on the mossy banks of the marsh while the scientists worked.

The years have gone by. Sherrod’s son has since completed graduate school in applied geosciences.

On a frigid, blustery day in December 2018, Sherrod revisited the site where he conducted much of his field work.

His team wanted to find the rate of sea level rise along the shore. An abrupt rise or decline in sea level would reveal if the fault had triggered a quake before.

By sampling sediment from the marsh to the beach berm, Sherrod and his research partner, Harvey Kelsey, developed a timeline of the ocean’s climb.

A few miles southeast across the white-capped waves of Admiralty Bay, Lake Hancock rises and falls with the tides. On an inactive fault, the sea would have risen at the same rate at both locations. But it didn’t. The marshy deposits are about a meter higher at Lake Hancock.

“Everything points to one thing,” Sherrod said, waving his hand across the inland sea. “This is an active fault.”

The evidence shows each lake rests on different free-floating jigsaw pieces of planetary crust, separated by the southern Whidbey Island fault. The height difference likely was caused by a 7.5 magnitude earthquake on the fault about 2,700 years ago, Sherrod said. He said he believes dramatic shifts from that quake also may be visible on the western edge of Camano Island.

Along the water at Cama Beach State Park, cabins on a bluff overlook Saratoga Passage, facing the general direction of Lake Hancock on Whidbey Island. The southern Whidbey Island fault divides the two.

The bluff, where the cabins now sit, could have jutted up in the most recent Whidbey fault quake, Sherrod said.

At the Brightwater treatment plant in Woodinville and at Crystal Lake in Maltby, the government researchers found telltale slopes of offset ground, known as scarps, indicative of a long-ago quake.

In the Puget Sound region, it takes a trained eye to recognize rocky outcrops and subtly raised ground as evidence of a fault. One of the best views of SWIF should be from Grand Avenue Park in Everett.

On a brilliant November day, Sherrod took in the panorama from the park’s bluff. Each peak of the Olympics stuck out with picturesque clarity in the distance. The fault, not so much.

“Here, we’re looking at one of the bigger faults in the region,” he said. “And what we’re looking at is water.”

The southern Whidbey fault is unlike more visible faults on the West Coast. The San Andreas in California, for example, left gaping scars in the Earth’s crust, at the surface.

SWIF ranges from 12 miles underground at its deepest to right at sea level in a few scattered spots, like Cama Beach, Holmes Harbor and Woodinville, according to Sherrod’s research.

The fault has at least three almost parallel strands within a 4- to 7-mile-wide band, stretching eastward from Vancouver Island.

Scientists are not sure how far east it goes. One model suggests it extends to about 30 miles east of Yakima. The fault’s length depends on whom you ask, Sherrod said.

During site visits in 2005, Sherrod’s team found evidence of four SWIF earthquakes in the past 30,000 years. The most recent hit roughly 2,700 years ago.

So when will the next one erupt?

Sherrod shrugged his shoulders. Scientists don’t know. They haven’t dug up enough history to estimate.

“I wish we did,” he said.