How Sound Travels Through Water

In the early 1940s Lineham reported a class of seismic waves that were observed only on coastal and island seismic stations. These are T-waves. T waves are seismically generated acoustic waves that propagate over great distances in the ocean sound channel. The “T” refers to tertiary, since these water-borne seismic waves travel slower than solid-earth P (primary) and S (secondary) waves, and thus arrive third on seismograph records.

Cartoon of an earthquake below the seafloor (red star) and the energy that radiates from the source location. When the energy intersects the seafloor it converts to sound in the water column and is eventually recorded by hydrophones.

The origin of the T-waves was first explained by Tolstoy and Ewing in 1950. They recognized that the phase arrival times recorded on land seismometers were consistent with ocean sound velocities and occurred only with events that permitted ocean propagation paths. What happens is that seismic events on or below the seafloor generate energy that is transmitted into the ocean water column. Some of this energy gets trapped within the SOFAR channel, and can be recorded by hydrophones suspended in the channel.

Speed of sound plotted against water depth. Note the sound velocity minimum near 1000 m water depth.

The SOFAR channel is a zone of low acoustic velocity. The speed of sound depends on the temperature of the water, its salinity, and the pressure (which is equivalent to depth below the sea surface). The speed of sound ranges between 1400 and 1570 m/sec (4593 and 5151 ft/sec). This is roughly 1.5 km/sec (just under 1 mile/sec) or about four times faster than sound travels through air. The graph below shows how sound speed varies as a function of water depth, and shows a sound speed minimum in the depth range of between 700 m and 1200 m below the sea surface. Sound waves that enter this region (SOFAR channel) bounce back and forth between the top and the bottom of the channel, and since the attenuation of seawater is very low, energy can travel very long distances, eventually coupling back into solid rock at the coastlines.

Figure shows the earthquake locations as white circles plotted on the bathymetry of the ridge crest. The earthquakes migrated down the slope of ridge indicating that a dike had been intruded along the ridge.

Hydrophones are now used to record T-waves from earthquakes generated throughout the ocean basins. In order to understand the nature of earthquakes at the Juan de Fuca spreading center, NOAA’s Pacific Marine Environmental Laboratory began monitoring the U. S. Navy’s SOSUS (Sound Surveillance System) hydrophones in the north Pacific on August 29, 1991. On June 26, 1993, recurrent, low-level (mb = 1.8 - 3.5) seismicity was recorded on the real-time system and was subsequently located on a small segment of the Juan de Fuca ridge, “CoAxial Segment”. NOAA, Canadian, and academic research vessels investigated the site of this first event. This effort was remarkably successful, and the nature and consequences of a magmatic seafloor spreading episode were thoroughly studied for the first time.

Red and orange areas mark the location of new lava flows that were mapped after seismic events were recorded in this region. Apparently magma within a dike that intruded along the ridge axis erupted on the seafloor to form new lava flows.

In February, 1996, another magmatic event was detected along the northernmost segment of the Gorda Ridge, leading to another coordinated and successful investigative effort. Because the magnitude of the events was so small, none of them were detected by land-based seismometers in the Pacific Northwest. The ability to monitor seafloor spreading events using hydro-acoustic methods is really enhancing our understanding of how mid-ocean ridges work.