How the UK’s Hurricane-Strength Storm of 1987 Became a Famous Weather Forecasting Failure

“Are you going to resign?” read the words, scrawled on a scrap of paper and pushed through the letterbox of a quiet London house.

Outside, reporters had gathered, demanding answers from one of Britain’s best-known weather forecasters, Michael Fish, who only the day before had reassured viewers that a hurricane wasn’t going to hit Britain.

Inside, his 11-year-old daughter, Nicola, bent down to collect the note.

“I remember just finding that really shocking,” she recalled recently, sitting on a sofa beside her father, in the same home where the message had landed 38 years ago. “Really, really shocking.”

It was Oct. 16, 1987. By dawn, southern Britain’s landscape was unrecognizable. Overnight, the country had been hit by its fiercest storm in more than 300 years. Winds howled at over 100 miles per hour, ripping through homes and countryside, felling an astonishing 15 million trees and tearing down telephone and electricity lines. At least 18 people were killed and damage reached over a billion pounds. The southeast of England, including London, bore the brunt of the devastation.

But it wasn’t only the destruction that shocked the nation. It was the fact that almost no one — not even Britain’s trusted weather forecasters — had seen it coming.

On the evening of Oct. 15, she had gone to bed — heavily pregnant — after watching the news. The bulletin, she remembered, had no warning of a storm about to hit Britain.

In the early hours, she woke with contractions. The wind rattled outside, though she thought little of it.

“It’s quite normal that time of year to be a bit windy at night,” she said. “So I went back to bed.”

A couple of hours later, as the contractions grew stronger, she and her husband John, decided it was time to go to the hospital. But as they prepared to leave, the power cut out. She recalled “fumbling around” in the dark, gathering belongings, before heading to the car, which was when she realized something was terribly wrong.

The journey that followed was harrowing. Branches and power cables lashed against their car, and fallen trees repeatedly blocked the road. With labor intensifying, Mrs. Pell found herself battling not just pain, but the storm itself.

They made it as far as the village of Ashurst, around five miles away from their home, and called the hospital from a telephone box. An ambulance was promised, but moments later came a callback — it was too dangerous for one to come. A helicopter would be sent instead, but then the line went dead. The village had lost power.

There was no choice but to turn back — on foot.

“I’d walk for a few minutes and have to stand still and have a contraction,” she said. “Then I was having to climb over these oak tree trunks and carry on along the road, and it was almost impossible to see where the road was because it was filled with so many trees.”

The fear was overwhelming. “I thought, oh, this is it, this is the end. This is how I’m going to go,” she said.

As the storm advanced inland, London began to darken.

For Tony Malins, then a manager with the electricity board, the chaos began with a phone call. As he reached for his bedside lamp, nothing happened.

He set out immediately for the control center in London, a 25 mile journey.

When he finally arrived, the picture was bleak. The entire southeast had lost power.

Circuits tripped and re-tripped as salt spray from the sea was getting into insulators, while winds buffeted transmission lines, causing them to clash and short circuit. Plastic bags and debris clung to the lines, causing additional problems.

If the storm had moved farther north, he said, the entire country could have lost power. It took transmission engineers up to three days to completely restore power to the southeast.

The Met Office’s annual report for 1987, in reviewing the forecast, said that notice of the storm had been posted as early as Sunday, Oct. 11, on a BBC farmers’ forecast. It warned of weather “becoming very windy late in the week.”

But as the week progressed, forecasts grew less certain. Early computer models that had hinted at a storm later dropped it altogether.

“They kind of lost it as we got closer to the time,” said Alex Deakin, a weather presenter and meteorologist at the Met Office. “The forecast basically missed it.”

Mr. Fish remembered that the models suggested the storm would pass south of England.

“It was showing that this thing would go over to France,” he said. “I don’t think there was any inkling whatsoever, until quite late, that this thing is going to take a different track.”

Gale warnings were issued to the public at 6:30 a.m. on Oct. 15, followed by more severe warnings a few hours later.

By that afternoon, however, conditions were calm across much of the country. During the BBC’s lunchtime news, Mr. Fish opened his weather bulletin with words that would later become infamous in British weather forecasting:

“Earlier on today apparently a woman rang the BBC and said she’d heard there was a hurricane on the way,” he said. “Well, if you are watching, don’t worry, there isn’t.”

In truth, Mr. Fish was correct. A hurricane was not heading toward Britain.

“You need temperatures in the sea to be 26 degrees Celsius or more,” Mr. Deakin said, or about 79 Fahrenheit. “If you’ve ever been bathing in the seas around the British Isles, you’ll know that the sea temperatures are never that warm.”

Some wind gusts, however, did reach hurricane strength. The Met Office recorded a peak gust of 115 miles per hour at Shoreham on the Sussex coast. And in parts of southeast England, where the greatest damage occurred, wind gusts of 80 miles per hour were recorded for three to four consecutive hours.

Mr. Fish became the press’s scapegoat.

A government inquiry into the storm later found that the fault did not lie with the weather presenters. They had simply delivered the forecasts provided to them by senior forecasters at the Met Office headquarters.

“In view of the briefing they received from the senior forecaster on this occasion they cannot be blamed in any way for the failure to the forecast the severity of the storm,” the report said. “The subsequent hostile reaction to them in some sections of the press was grossly unfair.”

In 1987, the Met Office’s weather computers could process about four million calculations a second — less powerful than a mobile phone is nowadays, Mr. Deakin said. This limited processing capacity meant that forecasts had lower resolution.

“When we’re making weather forecasts, we’re dividing the atmosphere up into boxes,” he said. “The smaller those boxes, the more accurate the weather forecast.”

At the time, the global model divided the atmosphere into grid boxes 150 kilometers wide. In more recent years, those boxes have shrunk, to 10 kilometers for global forecasts and as small as 1.5 kilometers over Britain, significantly improving accuracy.

The day before the storm, at the Met Office’s then headquarters in Bracknell, meteorologist Phil Garner noticed something strange. His job was to monitor incoming reports from ships and observation sites and correct any anomalies before they were processed for the forecasts.

The pressure readings he was seeing were “far too large that the machine could handle,” he said. “From memory, I think it was falling 10 millibars an hour or greater. Now that’s explosive cyclogenesis. That was a storm happening.”

The phenomenon, also known as a “weather bomb,” occurs when a storm’s central pressure drops by at least 24 millibars in 24 hours. But in 1987, the Met Office computers were too limited to detect it.

“The machine had no way of realizing that was happening,” Mr. Garner said.

Unable to input the real pressure fall of the storm, Mr. Garner resorted to approximations.

“It had to be a sort of by guess or by God type of thing,” he said. The storm became a turning point for British meteorology, driving decades of research, investment and improvement, from the introduction of satellite data to developing ensemble forecasts, which generate multiple scenarios from slightly different starting simulations of the atmosphere, to show both what might happen and how likely each outcome is.

Modern models can now recreate the atmosphere every six hours, going back more than 70 years, using archived observations. Scientists have therefore been able to rerun past weather events, with today’s technology.

The European Center for Medium-Range Weather Forecasts reran an ensemble of 50 simulations for the 1987 storm, starting two-and-a-half days before it hit.

Even today, Mr. Mylne said the forecast would remain uncertain. Its rapid intensification was very sensitive to various components coming together in an exact way, so small uncertainties in the starting state would likely lead to large variations in the forecast.

But these days, he said, “We would manage the difficulties and communicate the forecast much more effectively.”

Warnings would use language that focuses on likelihood and impact, changes the Met Office introduced in the wake of the 1987 storm. For better awareness, the storm would likely be named too, a step some countries have begun taking to help focus public attention on a specific weather threat.

If the storm was to strike again today, Mr. Deakin said, it would probably bring an additional layer of risk, because of climate change.

“It would definitely drop more rain because a warmer climate holds more moisture,” he said. “So there would be more flooding from the same system if it happened again.”

The forecast, and Mr. Fish’s role in it, has lived on in British memory for four decades.

His name remains forever linked to that extraordinary night. The now infamous “hurricane” clip even featured in the opening ceremony for the 2012 London Olympics. But time has softened the memory of those painful first days.

“The British public are wonderful actually,” he said. They are highly sympathetic and appreciate you do a difficult job.”

And for Mrs. Pell, the memories are far more personal. After walking with her husband, John, through the storm for nearly two hours, she gave birth safely at home, with the help of a nearby doctor.

Their baby girl was given a fitting middle name: Gayle.

Audio produced by Adrienne Hurst.