The D-Day weather forecast was flawed, but the Allies won and the war changed meteorology forever

A memorial honoring a World War II hero will be unveiled in Atlantic City on Thursday, the 80th anniversary of D-Day, and Marine Corps veteran Jim Eberwine, one of the principals involved, was understandably anxious on Wednesday.

Thunderstorms are possible, and more than most people, Eberwine, a retired National Weather Service meteorologist, is aware that he may not know for certain whether a thunderstorm is in the area when the event, which has been years in the planning and is due to get underway at noon, until he hears the rumbles or sees the flashes.

“Now I know how those forecasters felt making the forecast for the D-Day invasion,” said Eberwine.

He’s also aware that forecasting is a radically different exercise these days, in large measure because of World War II and the technological revolution it spawned, said Jordan Gerth, scientist with the weather service’s Office of Observations.

The prediction business still has serious weaknesses; for example, the timing of thunderstorms, and getting a handle of sky cover. For evidence, check out some of the sky forecasts for the April eclipse, which didn’t turn out as some had expected.

“I was in Dallas, a meteorologist, an expert in this, and I wasn’t sure it was going to clear two hours before,” he recalled.

Caveats aside, said Gerth, meteorologists now have access to high-speed computer models that ingest vast amounts of surface, upper-air, and satellite data. By contrast, the forecasters in the 1940s were working with cave drawings.

Today, he said, “We probably would have been able to provide forecasts five days out.”

In those days, forecasting was more labor-intensive. Three teams of U.S. and British meteorologists, including one who became a fixture on local television in Philadelphia, spent six months working on a forecast that underwent a radical change on the eve of the invasion. And in the end, the final version didn’t work out all that well.

“Weather forecasts in the 1940s just were not very good,” said Gerth.

For one thing, the Western world didn’t learn about the upper-air jet stream that moves weather systems until World War II pilots encountered it on bombing raids. (It had been discovered in the 1920s by the Japanese, who attempted, unsuccessfully, to weaponize it.)

Observations, so plentiful today, were limited to “ground-based weather reports, only from Allied sources,” said Gerth.

The forecasters were under tremendous pressure. Allied Commander Gen. Dwight D. Eisenhower wanted a 10-day outlook for Normandy for the days before and during the invasion. He was looking for clear skies, light winds, and calm seas.

Given that the operation involved 160,000 Allied troops, Eisenhower wanted a certain precision.

At the time, no one was even attempting 10-day forecasts.

The U.S. meteorologists examined charts showing how the weather behaved during the projected invasion period in past for clues for what might happen in 1944, what is known as the “analog method.”

Among them was Air Force veteran Francis Davis, who later became dean of science at Drexel University and one of the nation’s TV weathercasters. On the recommendation of the head of the American Meteorological Society, who also was on the D-Day team, shortly after the war Davis was recruited by Walter Annenberg, who owned The Inquirer and Channel 6. Davis became the station’s on-air weathercaster.

The story has been told and retold by a variety of sources through the years, pointed out the late Swedish atmospheric scientist Anderss Perrson in a paper published four years ago in the Bulletin of the American Meteorological Society.

Combing documents and examining phone transcripts, he concluded that some of the enduring perceptions about the high quality of the forecasts are questionable.

Eisenhower had his mind set on June 5, one of the three days in June with enough moonlight to provide visibilities for paratroopers and pilots during the amphibious assault, however a storm threat delayed it at the last minute.

Analyzing the available changes in barometric pressure, the meteorologists said they expected a “break” between storms on June 6, and the invasion was on.

The winds ended up being stronger than forecast, said Perrson. “Many soldiers suffered dreadfully from sea sickness,” he wrote. “Some soldiers resorted to vomiting into their helmets. They were thoroughly exhausted by the time they reached the beaches.”

With strong winds blowing onshore piling up water on the beaches, “the advantage of invading at low tide was lost”

Perrson said that on invasion day, “there had been a marginal break in the clouds and visibility, but there was no respite with the wind.”

The weather systems did not behave as predicted, he said, and as it turned out, the forecast was “right for the wrong reasons.”

Still, Eisenhower evidently was satisfied.

It has been reported that while riding to the Capitol for his inauguration, President-elect John F. Kennedy asked Eisenhower, his predecessor, why the invasion had succeeded. Eisenhower allegedly responded: “Because we had better meteorologists than the Germans!”

Eighty years later, meteorologists have a formidable arsenal of technology that would have been the stuff of futuristic hallucinations in 1944, but it was the imperatives of war that drove the development of meteorology, said Gerth, and the science benefitted further from the demands of the Cold War and Space exploration.

Still, for all the computer power and satellite advances, the atmosphere is imperfectly observed, that remains a handicap to computer forecast models.

Gerth said if that D-Day forecast were made today and involved thunderstorms, meteorologists couldn’t predict precisely what time and where they might occur.

Thunderstorms are likely throughout the region, including Atlantic City, the venue for the 90-minute ceremony dedicating a monument to local native Bernie Friedenberg, an Army medic who served on Omaha Beach during D-Day.

They event is scheduled to begin at noon.

The precise thunderstorm schedule was unavailable.