The Windsor Hum: The Low Rumble That Crossed a Border and Baffled Two Countries

About 7 min read

On some nights in Windsor, Ontario, the walls seem to breathe. Residents describe it as a low diesel engine idling somewhere just out of sight, a deep throb that you feel in your chest as much as hear with your ears. It rattles windows, sets dishes humming in their cupboards, and, for the people most sensitive to it, makes sleep impossible. It comes and goes without warning, sometimes absent for weeks, sometimes present for days at a time. It has been reported since early 2011, and it has been investigated by two national governments, two universities, and one of the most careful acoustic teams in Canada. The sound is real. Its source, the scientists concluded, is almost certainly a blast furnace on an island across the river in the United States. Almost certainly. That small gap between certainty and proof is where the Windsor Hum still lives.

Table of Contents

A city that could not sleep

Windsor sits at the southwestern tip of Ontario, directly across the Detroit River from Detroit, Michigan. It is an industrial border city, home to car plants, foundries, and the constant low thrum of manufacturing. Its residents are used to noise. What they began reporting in early 2011 was different: an intermittent, low-frequency rumbling that seemed to come from everywhere and nowhere, most noticeable late at night and in the early hours of the morning, when the rest of the city went quiet.

The reports shared a pattern. The sound was described as a low drone or rumble, sometimes compared to a truck idling outside the house or a distant generator. Some people heard it; others felt it as a vibration in the floor or in their bodies. It was worst in the city’s west end, the neighbourhoods closest to the river and to the industrial island on the American side. And it was maddeningly inconsistent. A person might hear it for three nights running and then not again for a month, which made it hard for officials to take the complaints seriously at first — and harder still for residents to prove that anything was happening at all.

The frustration was compounded by the fact that the hum did not respect the border. Whatever was making the sound appeared to be on the American side of the river, in Michigan, while the people suffering from it were in Canada. That meant the problem belonged to two countries, two sets of regulators, and two legal systems, neither of which could simply walk onto the other’s territory and start measuring.

The island across the river

At the centre of the story stands Zug Island, a heavily industrialised island in the Detroit River, just south of Detroit. The island began as marshland. Samuel Zug, a bookkeeper and furniture maker who went into real estate and politics, once imagined building a grand mansion there. Instead, in the early 1900s, it became an industrial hub, and today it hosts a steel mill complex, including the blast furnace operations of U.S. Steel. It is a guarded, smoke-and-steam-belching piece of land that the public cannot freely enter, and it sits almost directly across the river from the Windsor neighbourhoods where the hum is heard most strongly.

Natural Resources Canada conducted an early study that confirmed the low-frequency sound was real and estimated its source to be in the vicinity of Zug Island. That was an important step: it established that the hum was not imaginary and not a mass delusion, and it pointed investigators in a specific direction. But confirming that a sound exists and finding its exact source are two different problems, and Zug Island’s restricted access meant researchers could not simply walk in with microphones and measure the machinery directly.

Two universities, one river

In January 2013, the Canadian Department of Foreign Affairs and International Trade contracted the University of Windsor and Western University (then the University of Western Ontario) to conduct a joint acoustic study. The project was led at Windsor by mechanical engineering professor Colin Novak and his Noise, Vibration and Harshness — Sound Quality research group. The work was split into two phases: first, to identify and characterise the acoustic signature of the hum using monitoring equipment deployed across the Windsor-Essex community; second, to use advanced source-identification techniques to locate where the sound was coming from, without ever setting foot on the island itself.

The team deployed Brüel & Kjær noise monitoring terminals across the city’s west end. Western University deployed two portable infrasound arrays in the Windsor and La Salle area from late February to early March 2013, the arrays running until 8 April. The University of Windsor’s Great Lakes Institute for Environmental Research even lent the team a research vessel, so that scientists could drift up and down the Detroit River with portable infrasound arrays, recording noise within the hum’s frequency range from the water itself.

It was painstaking work, and the hum did not cooperate. For the duration of the project, it manifested on only a handful of days, which made identifying the source difficult. The infrasound arrays, for their part, could not geolocate a definite source: the Western University team reported that it found no common signals at both arrays that could plausibly be associated with the hum, and was unable to pin down a location using cross-bearings. Conclusive evidence from the river-based system was not achieved either, because the hum was not present on the days that equipment was deployed on the water.

What the Windsor team found

The University of Windsor’s work, however, produced a result. Six months of fieldwork followed by four months of analysis revealed a very high amplitude, low-frequency tone that matched the varied reports from the community. The researchers characterised the sound’s acoustic fingerprint and compared it against the possible sources capable of producing it. Their conclusion, published in the final report, was clear: the Windsor Hum does exist, and it has both qualitative and quantitative characteristics that point to the likely source being the blast furnace operations on Zug Island.

The Natural Resources Canada study had earlier identified a monotone signal correlated with hum reports by the public, with a dominant frequency of approximately 35 Hz — deep in the low-frequency range, at the very bottom edge of human hearing, where sound is as much felt as heard. The Windsor team’s findings were consistent with that. The report also noted that the hum’s periodic nature coincided with the observed irregular operations of the steelmaking operations on the island, and that the sound was audible to a concentrated portion of Windsor’s west-end community and nearby residential areas.

In May 2014, the Canadian government released the two-part report at a press event beside the Detroit River, overlooking the industrial skyline. The Member of Parliament who led the announcement stated that while the Western University report was inconclusive, the University of Windsor team had been able to confirm its findings scientifically. Novak himself later summarised the work plainly: based on all the data collected, the team determined there was no danger from the sound other than that it was very annoying, and that for people highly affected, it could disturb sleep.

Why a low sound is so hard to pin down

The Windsor Hum is a case study in why low-frequency sound is so difficult to investigate. Sounds at around 35 Hz have wavelengths of roughly ten metres, and they pass through walls, windows, and the ground with ease. They do not behave like ordinary noise. They bend around obstacles, travel long distances with little loss of energy, and can seem to come from everywhere at once. A person standing in a room may hear the hum loudly in one corner and not at all in another, because the sound waves reflect off the walls and create standing waves — pockets of loud and quiet within the same space. This is one reason why some residents heard the hum clearly while their neighbours heard nothing, and why the reports were so easy to dismiss at first.

It is also why the hum became a magnet for other explanations. Over the years, the Windsor Hum has been blamed on secret military projects, on the HAARP research programme in Alaska, on underground tunnelling, and on electromagnetic waves. None of these explanations has any supporting evidence, and the blast furnace on Zug Island — a huge, intermittently operated piece of industrial machinery sitting almost directly across the river from the affected neighbourhoods — remains the only candidate with a measured acoustic fingerprint to match.

The hum that would not go away

What makes the Windsor Hum remarkable is not that it is unexplained — by 2014, science had a good answer — but that a good answer was not enough to end it. The hum continued to be reported after the study. The sound’s source, if it is the blast furnace, is in another country, and the Canadian government cannot order a Michigan steel mill to change its operations. Mitigation has meant years of research and diplomatic collaboration between Canadian and American authorities, and the hum remains, for the people who hear it, an unresolved intrusion into their homes.

There is a larger lesson in the Windsor Hum, and it is the reason it belongs on any list of modern mysteries. The world is full of sounds that science can measure but cannot quite place, and the gap between measurement and proof is where rumour flourishes. The Windsor Hum was measured, characterised, and traced to a probable source by careful, patient science. It was also, for the people who lived with it, a sound in the dark that nobody could make stop. Both of those things are true at the same time, and that is what makes it a mystery worth remembering.

What is documented, and what is not

Documented facts: Since early 2011, residents of Windsor, Ontario, have reported an intermittent low-frequency noise, heard most strongly in the city’s west end. A Natural Resources Canada study confirmed the sound’s existence and estimated its source to be in the vicinity of Zug Island, a highly industrialised island on the US side of the Detroit River, identifying a monotone signal of approximately 35 Hz correlated with public reports. In January 2013, the Department of Foreign Affairs and International Trade contracted the University of Windsor and Western University to conduct a joint acoustic study; the Windsor team, led by Professor Colin Novak, deployed noise monitoring terminals across the west end and used a research vessel on the Detroit River. The final report, released in May 2014, concluded that the hum exists and that its likely source is the blast furnace operations on Zug Island. The Western University infrasound-array study was inconclusive and could not geolocate a definite source. The hum is not considered dangerous, though it can disturb sleep.

Unverified or disputed claims: No direct measurement has ever been made inside the Zug Island steelworks, because researchers had no access to the site, so the blast-furnace explanation rests on acoustic fingerprinting and correlation rather than on a measurement taken at the source itself. Claims that the hum is caused by secret military projects, the HAARP programme, underground tunnelling, or electromagnetic weapons have no supporting evidence and are contradicted by the measured acoustic signature, which matches industrial machinery. It is not established that every report attributed to the Windsor Hum has the same source; low-frequency sounds from different origins can sound alike.