A “Perfect” Tsunami

When an earthquake prompts tsunami evacuations, but no big wave, scientists see an opportunity.

By Yong Wei, for JISAO Magazine

DART in the open ocean
One of PMEL’s DART buoys floating in the open ocean.

When people think of tsunamis they often think of killer waves – a factual and life-saving connection.

The world remembers destructive tsunamis spawned by powerful earthquakes as if they happened yesterday. The 2004 Boxing Day tsunami in the Indian Ocean, the 2010 tsunami in Chile, and the 2011 East Japan event are the most recent examples. Altogether, these tsunamis claimed more than 250,000 lives and caused trillions of dollars in damages.

To enhance tsunami warning and forecasting capabilities, scientists drill deep to learn not only from the devastating tsunamis, but the unobtrusive ones as well.

Known to scientists as a “perfect” tsunami, these events, spawned by powerful earthquakes, pose little tsunami impact to coastal communities but are highly valuable in advancing our knowledge of future events in the same rupture area. They help us “debug” the current forecast system.

The Mw 7.8 earthquake/tsunami on July 22, 2020, was such a “perfect” event – a fault rupture offshore of Alaska Peninsula that led to less than one foot of tsunami yielding no damages along the coastlines nearby.

The earthquake occurred at 06:12:44 UTC according to the U.S. Geological Survey report. The epicenter at 55.068°N 158.554°W, was only 60 miles south of Perryville, a small borough of around 100 inhabitants, and more than 500 miles southwest of Anchorage (Figure 1a). A tsunami warning was issued for some parts of the Alaska Peninsula, the Aleutian Islands, and south Alaska. Residents of those areas, a stretch of more than 500 miles along the coast, were warned to head for higher ground and leave the designated danger zones.

IT WAS APPROACHING MIDNIGHT in Seattle when the strong quake nucleated off the coast of Alaska. A handful of CICOES/PMEL tsunami researchers monitored the wave fluctuation following the earthquake via the live data stream transmitted from DART (Deep-ocean Assessment and Reporting of Tsunamis), a global network of more than 60 PMEL-patented deep-water devices specifically designed to record tsunami water level with a bottom pressure recorder anchored to a nearby surface buoy (Figure 1b).

Tsunami figures and charts
Figure 1. Summary of the July 22, 2020 Mw 7.8 Alaska Peninsula earthquake and tsunami. (a) co-seismic deformation computed from the CMT-based DART-derived tsunami source of the earthquake; (b) max tsunami wave amplitude computed using the CMTbased DART-derived tsunami source of the earthquake; (c) model results vs observations at five nearby DART buoys and the tide gauge at Sand Point.

In less than one and a half hours, the tsunami waves peaked at five DART stations offshore spanning ~1,000 miles along the Aleutians and Alaskan Trenches. Somewhat surprisingly all heights appeared to be less than 1 cm, meaning a consequential, far-reaching tsunami would not be expected. The tide station at Sand Point harbor (Figure 1b), 80 miles northwest of the epicenter, registered up to 25 cm tsunami waves. As a result, the National Tsunami Warning Center canceled the warning about two hours later, after the threat had passed.

Indeed, this was a seemingly “perfect” tsunami producing a handful of data allowing us to unfold what we know and what we have missed.

SO WHAT DO WE KNOW? First of all, the NCTR/CICOES realtime modeling valuation tuned from the DART measurements predicted small waves in the near field. These predictions were further confirmed by a rapid model computation using the USGS’s finite-fault solution obtained from seismic data.

However, this solution predicted some misfits – larger wave amplitudes and earlier arrival – at all five DART buoys. Utilizing the new Graphic Processor Unit (GPU) model computation capability, NCTR/CICOES scientists later carried out a quick finite-fault tsunami source, driven by the USGS W-Phase solution, that offered a more reconcilable interpretation between tsunami (DART) and seismic observations (Figure 1b and 1c).

Here’s what we know about the tsunami so far:

  1. Large, shallow slips up to 4.3 m ruptured only a small portion (~600 km2) dipping at 20º along the subducting Pacific plate underneath the North American Plate. The energy released from the small rupture area was probably dissipated quickly when propagating away from the source.
  2. There were possibly high-slip ruptures in the deeper part of the Pacific Plate that contributed major energy to the seismicity but generated much smaller, longer tsunami waves.
  3. Multiple uninhabited islands blocked the tsunami energy from direct impact on coastal communities in the Alaska Peninsula.
  4. There are concerns whether this powerful earthquake has ruptured into the Shumagin Gap (Figure 1a), an area that has been considered immune from large earthquakes due to constant release of fault pressure.

“It kind of opens the door on what types of earthquakes could occur in that region,” and “what kind of tsunamis we should plan for,” said Michael West, the state seismologist at the Alaska Earthquake Center in an interview with Anchorage Daily News.

WHAT HAVE WE MISSED? Murphy’s law found its way to play a role in what should have been an opportunity to test the latest DART 4th generation (4G) model. The 4G technique has an exceptional ability to separate tsunami waveforms from the background seismic noise with its highfrequency (1 Hz) sampling rate (reporting at every 15 sec).

The DART 46403, located less than 200 miles south of the epicenter (Figure 1c), is equipped with this technology, however, the 4G capabilities were not engaged during this event. Instead it performed as a 2nd generation buoy with the tsunami signals entirely overwhelmed by the seismic noises due to its proximity to the epicenter. A study of the later model revealed the tsunami wave might have peaked at 2 cm about 20 minutes after the earthquake on DART 46403 (Figure 1c).

From this perspective, this “perfect” tsunami worked as a debugger of the buoy network and enabled an important correction to the forecasting system, making it better prepared for a future event.

Someday, when the next big one hits, all of us will surely benefit from these little perfect tsunamis spawned by powerful quakes.

2020 JISAO Magazine