The Bloop

Summer 1997. Somewhere in the remote South Pacific, something made a sound loud enough to be picked up by hydrophones more than 3,000 miles away.

NOAA's Equatorial Pacific Ocean autonomous hydrophone array — repurposed from Cold War-era US Navy SOSUS submarine-detection infrastructure after 1991 — caught it on multiple sensors simultaneously. Analysts named it the Bloop. The acoustic signature was a rapidly rising frequency lasting roughly a minute, which put it in the same general shape as a biological vocalization. The amplitude did not. It was louder than any known cetacean call, including the loudest blue whale signals ever recorded. NOAA's own assessment was consistent across the unidentified period: no known marine animal produces sound at that scale.

What happened

For fifteen years, the Bloop sat in NOAA's public catalogue of unidentified hydrophone events alongside other named mysteries — Julia, Upsweep, Train, Whistle, Slow Down. The catalogue is a strange document. It reads like a field guide to sounds the ocean makes when nobody is looking, filed under names that sound like they were assigned by someone working a very long night shift.

The Bloop's source was localized to roughly 50°S 100°W — a remote patch of the equatorial Pacific. That coordinate sits in the general oceanic neighborhood of R'lyeh, the fictional sunken city from H.P. Lovecraft's fiction, which the internet noticed immediately and enthusiastically. NOAA did not find this relevant. The agency's position throughout the unidentified period was that biological origin was implausible on amplitude grounds alone, full stop.

The evidence

By 2005, NOAA had recorded confirmed icequake events — large icebergs cracking off the Antarctic ice shelf — from the Scotia Sea and Bransfield Strait. When researchers ran the spectrograms of those events against the 1997 Bloop recording, the match was close: frequency profile, duration, and amplitude all lined up.

In November 2012, NOAA PMEL's Christopher Fox told Wired that the Bloop was a non-tectonic icequake — a large fracture event consistent with Antarctic ice-shelf dynamics. Fox noted that NOAA had been confident in the icequake hypothesis since approximately 2008 but hadn't made a clean public statement until then. Four years of quiet confidence before the announcement. The science moved faster than the press release.

The icequake attribution fits. Non-tectonic icequakes can produce exactly the kind of rising-frequency, high-amplitude, broadband signal the Bloop showed. The geographic proximity to Antarctic ice activity is consistent. The spectrogram comparison is the load-bearing evidence here — not a theory, a visual match against confirmed events.

What's still open

The Bloop itself is resolved. The icequake attribution is solid, the spectrogram comparison holds, and NOAA closed the file.

Everything else in that catalogue is not resolved. Julia, Upsweep, Train, Whistle, and Slow Down remain unattributed. Some have proposed explanations; none have a clean spectrogram match like the Bloop eventually got. The array is still listening. The catalogue is still growing.

The Bloop's fifteen-year residency in the unidentified column is also worth sitting with. The resolution came from patient comparison against later-recorded confirmed events — not from a new investigation of the original signal, but from the ocean eventually producing something similar in a place where scientists could verify it. That's how a lot of geophysical mysteries get solved: not by cracking the original case, but by waiting for the phenomenon to repeat itself somewhere more convenient.