The Right Place, at the Right Time, with the Right Equipment
On July 29, 2021 at 06:15 UTC (July 28, 2021 at 23:15 PDT, local time) a magnitude 8.2 earthquake struck offshore the Alaska Peninsula near Perryville, AK at a depth of ~32 km below the seafloor. This earthquake was the largest event in the US since 1965 and the largest in the world since the Fiji earthquake in 2018. It caused widespread tsunami warnings across the Pacific rim, with evacuations on the Homer Spit in Alaska. Thankfully, no major damage was caused by the resulting small tsunami. All warnings were downgraded and eventually cancelled.
In another region of the Pacific Ocean, ~1600 km away, the R/V Marcus G. Langseth was collecting active source seismic data to assess the tectonics and hazards associated with the Queen Charlotte Fault. To collect this data, the R/V Langseth was towing a 15 km long acoustic receiving device (called a streamer) to record acoustic waves generated by the ship to image the sub-seafloor geology. The scientific team also deployed a series of ocean bottom seismometers (OBS) to record the acoustic energy from the ship and any seismic events, such as the Perryville earthquake (see previous blog post by Maureen Walton).
At approximately 06:34 UTC (23:34 PDT) a massive spike in the noise on the R/V Langseth hydrophone streamer caused onboard scientists (including myself) and technicians to take notice. What just happened? Why was the noise level so high all of a sudden? Is something wrong with the equipment? After some quick discussion and examination of the data, we noted the noise was traveling down the streamer, from the ship to the tail buoy over 15 km behind us. We also noted that the noise was a linear arrival traveling at a constant speed along the steamer (straight diagonal lines in image below). After ~40 seconds the noise was mostly gone, rapidly dissipating, so it didn’t seem like an instrument issue.
We quickly noted this event looked similar to events caused by earthquakes seen on other scientific expeditions in Chile (by myself in 2017) and along the Aleutians near Dutch Harbor, AK (by technician Gilles Guérin in 2020). So, Klay Curtis (technician on board the R/V Langseth), Laz Garza (WWU graduate student), and I searched the USGS earthquake website to find recent earthquakes that could have introduced energy into the water column, and we found information on a massive earthquake across the Pacific. The timing was right for an acoustic arrival traveling at ~1500 m/s (sound speed in sea water), so this could be our noise source!
A normal “shot gather” showing the data we collect from the 15 km streamer.
A “shot gather” overwhelmed by the earthquake energy! Note the earthquake noise appears as linear features in the data window.
After loading the data collected by the streamer, I explored finding the azimuth of the arrival to verify the source of the noise. We recorded the acoustic wave propagation along the entire 15 km streamer, which allows us to calculate the velocity of the traveling wave. The apparent velocity of the noise was ~1820 m/s, much faster than the 1500 m/s speed of sound in sea water. However, this noise was not generated directly in front of our ship, instead the energy was traveling obliquely to our heading, making the velocity appear much higher. Using the apparent velocity, I did a little trigonometry (thank you high school trig) to determine the difference in our heading (326°) and the azimuth of the earthquake arrival (289° based on USGS event localization). I calculated an azimuth difference of ~35° based on the streamer data, very close to the 37° in reality. (For you nerds out there, this is without accounting for the streamer feathering with was ~1.2° to the starboard, which would bring our azimuth difference to within 1°! Pretty good for a scratch pad of paper, a little trigonometry, and an opportune acoustic receiver).
8.2 Perryville, AK earthquake (pin) and our approximate location offshore Haida Gwaii in the southeast. The red line shows the path the T-phase traveled to reach us over 1600 km away.
In the end, our working hypothesis is that our hydrophone streamer recorded a T-phase, or an acoustic wave generated by an earthquake that travels at the speed of sound through water. We think this T-phase was generated by the 8.2 Perryville, AK earthquake. The seismic energy propagated from the ~32 km deep rupture location offshore the Alaska Peninsula to the seafloor and down the continental slope. There, the seismic energy interacted with the SOFAR channel (I’ll let you read more about that without me), which converted the energy into acoustic waves in a process called “downslope conversion.” This interaction “trapped” the acoustic waves in the water column and helped propagate them, without losing too much energy, across the Pacific Ocean. Approximately 19 minutes later, the R/V Langseth recorded this event on the hydrophone streamer to the surprise (and eventual exhilaration) of the onboard crew. Without the T-phase and SOFAR channel interaction, we likely would not have seen this event on the streamer, even given the huge amounts of energy released during the earthquake rupture (but our deployed OBS would definitely record the arrival). It was a fortuitous set of circumstances that we were at the right place, at the right time, with the right equipment to capture this unique occurrence.
All in a night’s work onboard the R/V Marcus G. Langseth! Special thanks to Klay Curtis for nerding out with me as we explored this interesting event.
Benjamin Phrampus, US Naval Research Laboratory research geophysicist and co-chief scientist aboard the R/V Marcus G. Langseth.