On 4 July 2003, beachgoers at Warren Dunes State Park, in the US state of Michigan, were enjoying America’s Independence Day holiday when a fast-moving line of thunderstorms blew in from Lake Michigan. They scurried for shelter, but the event passed so quickly it didn’t appear that their holiday was ruined.
“In 15 minutes it was gone,” says civil engineer Alvaro Linares of the University of Wisconsin, Madison.
But when swimmers re-entered the water, rip currents appeared seemingly from nowhere, pulling eight people out into the lake, where seven drowned.
What these people had encountered, Linares says, was a meteotsunami — an aquatic hazard of which few people, including scientists, were aware of until recently.
Few scientists have researched the phenomenon. May of those who have gathered recently at the annual American Geophysical Union Ocean Sciences meeting, held in Portland, Oregon, US, to compare notes.
Conventional tsunamis are caused by underwater processes such as earthquakes and submarine landslides. Meteotsunamis, as the name indicates, are caused by weather. But while the catalysts are different, the effects are not.
“The wave characteristics are very similar,” says Eric Anderson of the Great Lakes Environmental Research Laboratory of the National Oceanic and Atmospheric Administration (NOAA) in Ann Arbor, Michigan.
To create a meteotsunami, what’s required is a combination of a strong, fast-moving storm and relatively shallow water. The sudden increase in winds along the storm front, possibly combined with changes in air pressure, starts the process by kicking up a tsunami-style wave that runs ahead of it. But the process would quickly fizzle out if the water was too deep, because in deep water, such waves propagate very quickly and would soon outrun the storm.
What’s needed to produce a meteotsunami is a water depth at which the storm’s speed and the wave’s speed match, allowing the wave to build as it and the storm move in tandem. “The storm puts all its energy into that wave,” Anderson says.
Furthermore, the wave can magnify even more when it hits shallower water or shoals. “That is when these become destructive,” Anderson says.
In 2004, for example, a storm front 300 kilometres wide sped across the East China Sea at a speed of 31 metres per second, 112 kilometres per hour, says Katsutoshi Fukuzawa of the University of Tokyo.
Water there is shallow, he adds, with depths mostly under 100 metres. This limits wave speed to about 30 metres per second — a near-perfect match to the storm’s. As a result, parts of the island of Kyushu were hit with a tsunami as big as 1.6-metres.
Not that meteotsunamis have to be that big to be dangerous. The one at Warren Dunes was probably no more than 30 centimeters, says Linares — small enough not even to be visible in the lake’s normal chop. {%recommended 4872%}
But unlike normal surf, meteotsunamis produce a sustained slosh that lasts several minutes between run-up and retreat. That means that even low-height waves carry a lot of water, creating the potential for strong rip currents when they withdraw. According to Linares’ models, these currents would have persisted for about an hour — plenty long enough to drag unwary swimmers far out into the lake, long after the storm had passed.
It’s also possible for meteotsunamis to become “detached” from the storm front that created them, striking shores far away. Researchers reviewing records in the Great Lakes have concluded that that is what happened when such a wave hit Chicago in 1954, killing 10 people.
“The wave came out of nowhere,” Anderson says. “It was a calm, sunny day.”
It’s not just Japan and America’s Great Lakes that have seen such events. In May 2017, a storm raced up the English Channel, kicking up a metre-high wave that swept beaches in The Netherlands as bystanders looked on with awe, says Ap van Dongeren of the Deltares research institute in Delft, The Netherlands.
Quirks of topography can magnify the effects of such tsunamis. On 13 June 2013, a group of spearfishermen in New Jersey were stunned when a surge of water threw them across a breakwater into the open ocean. A few minutes later, another surge threw them back where they’d come from. And that came from a meteotsunami that measured at well less than a metre on local tide gauges, says Gregory Dusek, a NOAA oceanographer at Camp Springs, Maryland.
Meteotsunamis have occurred on all inhabited continents, including one that hit the port of Fremantle, near the Australian city of Perth, in 2014, causing a ship to break free from its moorings and crash into a railroad bridge in 2014, Sarath Wijeratne of the University of Western Australia reported in a conference abstract. In fact, Wijeratne concluded, a look back at historical water level records indicates that Western Australia may have seen more than 15 such events each year between 2008 and 2016.
Other researchers are also finding these events to be surprisingly frequent. By studying tide gauge records back to 1996, Dusek has concluded that they occur on America’s eastern seaboard at a rate of 23 per year — though most are small enough nobody would ever notice. In Holland, Van Dongeren says that a quick check of historical tide gauge records revealed at least three such events in the past decade that had gone unnoticed because they happened at low tide. “They’re not that rare,” he says.
Fukuzawa says that Japan saw 37 meteotsunamis exceeding one metre from 1961 to 2005.
Furthermore, bigger ones are possible. In June 2014, Croatia was hit by a two-to-three metre tsunami sweeping in from the Adriatic Sea, says Clea Denamiel, of the Croatian Institute of Oceanography and Fisheries.
But the mother of all meteotsunamis came in 1978, when Vela Luka, at the southern end of Croatia’s scenic Dalmatian coast, was smashed by a meteotsunami measuring a full six metres, with giant waves surging and retreating about every 17 minutes, just as might have occurred in the aftermath of a large offshore earthquake.
As of now, scientists don’t know enough about meteotsunamis to be able to predict them, though efforts are under way to create models that can do just that. But as they dig back through old records, they are increasingly realising that meteotsunamis might have been with us for a long time.
Or as Linares puts it with typical scientific understatement, “meteotsunamis are a beach hazard that has been overlooked”.
