With transient killer whales sighted in the vicinity yesterday and the whereabouts of the endangered southern residents unknown, we were surprised and concerned to hear active mid-frequency sonar and distorted underwater voices this evening in Haro Strait. Thankfully, the broken hydrophone at Lime Kiln State Park had been repaired just hours earlier by Beam Reach students and staff conducting a service project at The Whale Museum‘s SeaSound laboratory that is housed within the light house. Eventually, Jeanne Hyde heard from the Coast Guard that we were likely listening to a U.S. Navy submarine. (4/9 update from Christopher Dunagan: the Navy confirms it was the USS San Francisco, a LA-class fast attack submarine which ran into a seamount near Guam in 2006. Since there was clearly damage to the sonar dome and it was replaced with the nose of decommissioned sub, I speculate that it was a higher priority of the Navy to test the sonar while still close to the ship yard than it was to go beyond the call of Naval duty to protect marine mammals.)
At 18:55 PST, Jeanne first reported someone “talking” into the Lime Kiln hydrophone of the Salish Sea Hydrophone Network. At 19:10, I started recording the Lime Kiln stream, capturing about 6 sequences of 5-10 distorted human words over about 11 minutes. The sequences seem to come every 1-3 minutes and begin with the same first few words and then vary in content and length. Most words had most of their power at 1.5-3.5kHz. About a minute and 15 seconds after the preceding word sequence, there was a 1.7-second tonal sonar ping at 2.9 kHz. At right is a screen grab from Audacity that shows the ping and the preceding word sequence. Here is a concatenation of the sequences and the ping: 090407_1910clips-lk_voice-sonar
Listening to both west side streams at around 19:15 it seemed that the voice and pings were louder at Lime Kiln, though they could be heard after a few second delay at the Orcasound hydrophone. By 21:00 it seemed the signals were louder at Orcasound than at Lime Kiln. This suggests that the Navy submarine was moving northward up Haro Strait. Since we have only a single hydrophone at each node and node-to-node timing isn’t synchronized in the streams, it’s not possible to verify the direction of travel using arrival time differences.
At 21:02 I made another recording from the Lime Kiln stream that contains two of the word sequences. Then I switched to recording the Orcasound stream. The voice was clearly audible for the next half hour and there were a few sonar pings interspersed. By 21:34 the voice was pretty faint and the last time it’s dectable in the sonogram (and by ear, very faintly) is around 10:05. However, at about 21:59 the final recording caught a couple of unique pings (3.7-3.8kHz, 0.8 second) and their echoes ( 090407_2150_os_375khz-sonar ). The two pings were spaced 16 seconds apart and had a noticable echo at 3.67 seconds, equivalent to about 5.5 km. Assuming the echo was off the eastern wall of Haro Strait bathymetry (about 6km from Orcasound), the two-way travel time suggests the submarine was probably pretty close to the San Juan Island side of Haro Strait.
Longer, unedited recordings are available at http://orcasound.net — including:
- Lime Kiln stream recordings (scroll down to date in Flash player)
- Orcasound stream recordings (scroll down to date in Flash player)
- Lime Kiln automated recordings
- Orcasound automated recordings (sonar pings)
- John Boyd’s compilation of sonar sounds (Quicktime movie format)
- OrcaGirl’s stream recordings (start times not apparent)
The acoustic experience continued on past midnight. At 00:24:30 on 4/8/09 I heard a series of beeps (back at ~2.9 kHz) and then their echoes at Orcasound. At 00:30-00:32 I again recorded the voice along with tones and their echoes on the Lime Kiln stream. Around 00:40-00:45 I heard and recorded what sounded like a conversation, rather than the separated single words of the previous sequences. At 00:48:30 there was a very loud ping (close to Lime Kiln?) and either an echo or a ping from a separate, more distance location. This was followed by many repeated word sequences. At 01:00 things got quiet, but then at 01:11:00 there was suddently another impressive single ping (sound clip: 090408_0038_lk_single-375khz-ping ). At 01:13, the slow voice started repeating sequences again. Another single ping came at ~01:22 as two cargo ships came abeam of Lime Kiln and increased the ambient noise levels. Another at 1:31:00 with a nice echo was followed by more conversing at 01:36. Then there was an extended sequence of sonar tones and verbal interjections that grew progressively fainter and then stopped ~01:45. Around 01:50 things picked up again with more conversation and then — most interestingly — what sounded like an exchange of longer tones, possibly between two sources, one near Lime Kiln and one substantially further away. After about 5 minutes of silence, there were two more loud pings with echoes at 02:00:45.
Recordings made for much of the following hour, but I think I missed two loud pings at ~02:53:00. More pings heard at: 03:11…