Seismic Data on Deck: Sounding for the Cascadia Megathrust Fault

A diagram showing the R/V Marcus G. Langseth‘s sound emitters and sound waves bouncing off of different layers below the seafloor. A streamer of hydrophones towed behind the ship listens for these echoes.

To help us in our search for the megathrust fault in the Cascadia subduction zone, the R/V Langseth is towing behind it a seismic sound source and a 12-km (~7.5 mile) streamer carrying hydrophones. With this equipment, we can collect a tremendous amount of data. But what are we actually collecting?

Every time we release sound waves from our source, we record on the streamer energy that has traveled into the earth, bounced off layers of rock from the seafloor to the mantle, and returned to our streamer at the sea surface. The Langseth’s long streamer has 960 channels that record the energy for 15 seconds at a sampling rate of 2 milliseconds. This is enough time for the sound waves to travel deep into the earth’s crust and return to the sea surface to be recorded by our seismic streamer before repeating the process. However, because our streamer is very sensitive to help listen for faint geological signals bouncing back, we record not only signals from the earth, but also a range of noise that can be carefully removed to help us see what’s beneath us.

Why do we need such a long seismic streamer if we’re just trying to look at what’s beneath us? There are two main reasons. Firstly, we can use the multiple recordings of the sound reflecting off each geological layer to help boost the reflection signal and reduce the amount of noise. Future work will focus on increasing the signal from geological features we’re interested in while reducing noise that can hide important information of the subsurface within it. Secondly, the long streamer helps us look at how the sound waves travel through earth’s crust, and this gives us velocity information. With both time and velocities in hand, we can then place geological layers in depth beneath the sea surface and allow for true geological interpretation of the data.

Example of what is recorded after a single pulse from the seismic sound source during our experiment. Mild filtering was applied to help remove random unwanted noise and help increase the signal from the seafloor and underlying layers. Additional processing will take this raw data and create an image of the geology beneath the seafloor. This will then allow us to find and map out the megathrust fault along the Pacific Northwest.

With the megathrust fault not only being located offshore of Cascadia but also kilometers beneath the seafloor, we need all the data we can get to tease out what the fault looks like.

Brian Boston is an associate research scientist at Columbia University’s Lamont-Doherty Earth Observatory.