In August 1991, scientists at the National Oceanic and Atmospheric Administration switched on a new way of listening to the ocean — and almost immediately heard something they could not explain. From a remote stretch of the South Pacific came a long train of tones, each one sliding upward in pitch like a siren heard through water. The sound was loud enough to be recorded across the entire Pacific Ocean. More than thirty years later, hydrophones are still hearing it. Its exact source has never been found.
The sound is called the Upsweep, one of a small family of unexplained ocean sounds NOAA recorded in the 1990s, alongside the Bloop and Julia. The Bloop, recorded in 1997, was eventually matched to icequakes — the cracking of Antarctic ice. The Upsweep has resisted that resolution: NOAA still lists it as unresolved, and the leading explanation, underwater volcanic activity, remains an inference, not a confirmed identification.
Listening to an Ocean
The sound was picked up by NOAA’s equatorial autonomous hydrophone arrays — networks of underwater microphones anchored in the deep ocean, designed to listen for low-level seismic activity and volcanic eruptions along the Pacific’s spreading centres. The arrays were part of the Sound Surveillance System, or SOSUS, a network originally built by the United States Navy to detect submarines and later adapted for oceanographic research by NOAA’s Pacific Marine Environmental Laboratory.
Sound travels far further in water than in air, and very low-frequency sound furthest of all — a loud enough signal can cross an entire ocean basin. The Upsweep is such a signal, recorded throughout the Pacific, which is one reason it drew attention so quickly.
The sound was already present when the Pacific Marine Environmental Laboratory began recording with SOSUS in August 1991. That means 1991 is the date of its first recording, not necessarily the date it began. It may have been sounding for years, decades, or far longer before anyone was listening.
What the Sound Looks Like
The Upsweep is not a single noise but a pattern. It consists of a long train of narrow-band upsweeping sounds, each lasting several seconds, repeating in sequence. On a spectrogram — a visual display of sound over time — it appears as a series of vertical sweeps rising from low to high frequency, each lasting roughly three seconds. Played back at twenty times its original speed, so that the very low frequencies become audible to human ears, it has an eerie, reverberating tone that has been compared to an ambulance siren.
The sound is seasonal, generally peaking in spring and autumn, but scientists are not sure why: it is unclear whether the source itself changes with the seasons, or whether seasonal changes in the ocean — temperature, currents, and how sound travels through water — simply carry the sound further at certain times of year.
The Upsweep’s volume has been declining since 1991: year by year it has grown fainter on NOAA’s arrays, though it can still be detected. No one knows whether the source is slowly quieting, moving, or whether ocean conditions carrying the sound have changed.
A Location in the Middle of Nowhere
By comparing the times at which the sound arrived at different hydrophones, scientists have roughly located its source: about 54 degrees south, 140 degrees west, in a remote region of the Pacific Ocean between New Zealand and approximately 2,500 miles due west of the southern tip of South America. That is open ocean, far from land, far from shipping lanes, and far from any settlement.
The location matters because it sits near an area of inferred volcanic seismicity — the foundation of the leading explanation. But the estimate is rough, the exact location unknown, and the sound has never been traced to a specific volcano, vent, or stretch of seafloor.
The Whale Theory — and Why It Was Dropped
When the Upsweep was first heard, researchers considered a biological origin. The most obvious candidate was the fin whale, a species known for powerful low-frequency calls. For a time, that explanation seemed plausible. By 1996, however, it had been dismissed.
The problem was variation — or rather, the lack of it. A whale changes its pitch, rhythm, and phrasing from call to call; the Upsweep does not. Its tones repeat with mechanical regularity, with too little variation to be biological, pointing toward a physical, geological process rather than a living one.
The Volcanic Explanation
The leading theory today attributes the Upsweep to underwater volcanic and seismic activity. Submarine volcanoes are common along the seafloor spreading centres — the rift zones where tectonic plates pull apart and new crust forms — that run through all of the Earth’s major ocean basins. These are exactly the regions the SOSUS arrays were set up to monitor.
Underwater volcanoes make a variety of sounds, from sharp cracks to booming explosions and low rumbles, and geologists are unsure what causes all of them. An ongoing process — seawater interacting with submarine lava flows, or gas released from hydrothermal vents — could plausibly produce a repeating, seasonal sound over decades, and the Upsweep’s location fits that picture.
But the fit is circumstantial: no eruption has ever been observed at the estimated location at the same time as the sound, and no hydrophone has been placed close enough to pinpoint the source. The volcanic explanation is the best available account, but it is unconfirmed, and NOAA still describes the origin as unresolved.
Why a Sound Can Stay Unexplained
Part of the reason it stays unexplained is geography: the source lies in one of the most remote stretches of ocean on Earth, and the arrays that hear it listen from thousands of kilometres away, able to estimate a direction but not a precise address. Closing that gap would require instruments much closer to the source — and no one has yet put them there. For now, the Upsweep remains what it was in August 1991: a rising tone from the deep, heard across an ocean, its maker unknown.
What Is Documented — and What Is Not
Documented: The Upsweep was present when NOAA’s Pacific Marine Environmental Laboratory began recording with SOSUS in August 1991. It consists of a long train of narrow-band upsweeping sounds, each lasting roughly three seconds, loud enough to be recorded throughout the Pacific. Its source is roughly located at 54°S 140°W, between New Zealand and about 2,500 miles west of the southern tip of South America. It peaks seasonally in spring and autumn, and its overall volume has been declining since 1991, though it can still be detected. A biological origin, initially attributed to fin whales, was dismissed by 1996 because the sound shows too little variation to be animal-made.
Not established: The exact source. Underwater volcanic activity is the leading explanation, but no specific volcano, vent, or eruption has been linked to it, and NOAA lists the origin as unresolved. Whether the seasonal peaks reflect changes in the source or in how sound travels through the ocean is also unknown.