The study, accepted for publication in the journal Weather ahead of the release of the feature film Pressure, which dramatises Stagg’s dilemma with Andrew Scott in the title role and Brendan Fraser as General Eisenhower, generated fifty separate forecasts of the June 1944 storm using machine-learning prediction techniques. Rather than producing a single deterministic answer, the researchers examined how confident the AI system was about the conditions that Stagg had to assess for 5 and 6 June. The results were strikingly mixed. For 5 June, the original invasion date, the model placed the storm in the correct position and correctly anticipated poor conditions over Normandy, in agreement with Stagg’s own reasoning. Yet for 6 June, the day the landings actually took place, the model made a consequential error, predicting that the storm would sweep away towards Norway far more quickly than it truly did and that the invasion day would be calmer and clearer than reality.

Professor Andrew Charlton-Perez, lead author of the study, framed the historical stakes in unambiguous terms. Stagg, he noted, held one of the hardest forecasting jobs in history: get the prediction wrong and the invasion could have failed, with consequences for the entire course of the Second World War. Some Allied meteorologists, after weeks of predominantly hot and settled weather, were insistent that conditions would remain fine for the invasion to proceed as planned on 5 June. The new research suggests that Stagg, had he been handed the AI forecast alongside his own charts, would likely have reached the same conclusion he did historically. The model pointed to a genuine risk of dangerous winds on 5 June, a signal strong enough to justify recommending the one-day postponement that ultimately allowed the landings to proceed under a brief lull in the weather.

The story of 6 June, however, reveals the uncomfortable limits of even the most advanced machine-learning systems. Eisenhower’s famous order, “OK, we’ll go,” rested on Stagg’s belief that conditions on 6 June would be fair enough to launch. In reality, soldiers landing on the Normandy coast encountered high winds and stormy seas more treacherous than the reassuring picture painted by the model. The AI system showed the storm clearing away from Britain too quickly, incorrectly suggesting that 6 June would be calm and bright. It also predicted clearer skies than actually occurred, compounding the false sense of security. In other words, a modern forecasting system trained on decades of global atmospheric data, with access to vastly more observations than Stagg ever had, would still have underestimated the very storm that made the D-Day landings so hazardous.

Perhaps the most technically revealing finding lies beneath the model’s overall confidence. Hidden within its aggregated forecast was a wind statistic that told a very different story: the model still assigned a 30 percent probability that winds on 6 June would exceed the safety limit for the landing craft. That significant residual risk was masked by the model’s reassuring average picture, illustrating a subtle challenge in how probabilistic ensemble forecasts communicate danger. An overall forecast that reads as benign can conceal a meaningful tail of adverse outcomes, and it takes a trained human forecaster, scrutinising the distribution rather than the mean, to extract that signal. Stagg himself, weighing conflicting advice from competing forecasting teams in the manner dramatised in the film, performed precisely this kind of synthesis with the tools of 1944: surface observations, limited upper-air soundings, and hard-won experience of Atlantic weather behaviour.

The methodology behind the study is as noteworthy as its conclusions. To judge how well the model performed, the researchers compared its forecasts against what actually happened over the Channel and the Normandy beaches, drawing on modern weather records from the Copernicus Climate Change Service, which reconstructs historical atmospheric conditions. Because weather records from 1944 were sparse, the team also turned to an unexpected archival treasure: newly rediscovered paper weather logs from the Faroe Islands, held in an archive at the Met Office. The Faroe Islands sit close to the path the storm followed as it passed Britain, so the readings recorded there in 1944 captured the storm at a critical point that the main observational networks missed. The comparison offered an independent check on the system’s position and intensity that would otherwise have been impossible.

Converting those handwritten logs into digital data allowed the researchers to assess the storm’s strength more closely, and the exercise produced a small but meaningful revision to history: the June 1944 storm may have been slightly stronger than current reconstructions suggest. This finding has implications beyond historical curiosity. Reconstructions of past weather underpin how scientists validate forecasting systems against real events, and even modest corrections to storm intensity can change how a model’s errors are interpreted. The researchers argue that digitising more historic paper logs from archives across Europe would allow storms like the one that threatened the D-Day landings to be tested with even greater precision in the future, enriching the observational record precisely in the oceanic and remote regions where modern reanalyses remain least constrained.

The wider lesson the authors draw speaks directly to the current enthusiasm for artificial intelligence in weather prediction. Machine-learning forecasts can process enormous volumes of data in seconds, producing global outlooks at computational speeds that traditional numerical models, which simulate atmospheric physics on supercomputers for hours at a time, cannot match. Yet the D-Day replay demonstrates that these systems can still make mistakes, particularly in unusual or extreme situations that sit outside the statistical patterns they learned from training data. A historical storm from a data-sparse era is an especially demanding test case, and the model’s errors on 6 June, its over-rapid clearing and its excessively optimistic skies, are exactly the kind of failure that could prove costly if trusted uncritically.

Charlton-Perez’s conclusion is that the human forecaster remains as essential today as in 1944. Having a trained meteorologist make sense of forecast data, interrogate the probabilities, and apply knowledge and experience to the final call is, in his words, just as critical now as it was for the D-Day landings. The study stands as a rare natural experiment: a chance to pit a state-of-the-art AI system against a decision that altered world history, using the same deadline pressure and far better tools. Stagg passed the test with paper charts and courage. The machine, for all its speed and scale, got half of it wrong, and a thirty percent hidden wind risk reminds us why someone still needs to read between the lines of every forecast before the order is given to go.

Subject of Research: Testing modern AI weather forecasting models against the historical D-Day storm forecast of June 1944.

Article Title: AI D-Day weather forecast falters under pressure

Article References: AI D-Day weather forecast falters under pressure. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: AI weather forecasting, D-Day, James Stagg, machine learning, Normandy invasion, storm forecasting, Weather journal, University of Reading, Copernicus Climate Change Service, ensemble forecasts, historical weather records, meteorology

The research, published in Advances in Atmospheric Sciences, focuses on a fundamental challenge in severe-weather prediction. Thunderstorms are driven by atmospheric processes that can change rapidly across very short distances. Warm, moist air may surge into one neighborhood while cooler, denser air spreads across another. Pressure can fall or rise as storm-scale circulations develop, but conventional weather stations are often separated by many kilometers. That spacing may be sufficient for monitoring broad weather patterns, yet it can miss the localized signals that determine where a storm strengthens, produces hail, or unleashes intense rainfall. A dense network of low-cost observations could provide weather models with a much more detailed picture of the atmosphere near the ground.

Many smartphones contain microelectromechanical-system barometers designed primarily to support functions such as altitude estimation, indoor navigation, and location services. These sensors measure atmospheric pressure electronically, generally detecting tiny changes through the movement of a microscopic mechanical component. Although an individual phone’s readings can be affected by temperature, device design, indoor conditions, or elevation, large numbers of measurements may reveal meaningful patterns when the data are carefully calibrated. The researchers explored whether these imperfect but abundant observations could be transformed into useful information for numerical weather prediction.

For their test, the team examined a severe hailstorm that struck Beijing on June 30, 2021. The researchers used anonymized pressure measurements collected, with users’ consent, through the Moji Weather mobile application. Before the observations could be used, the raw readings had to be corrected for errors and inconsistencies. The team applied machine-learning methods to process the measurements, accounting for differences among devices and attempting to distinguish genuine atmospheric signals from noise caused by buildings, sensor behavior, and other local effects. The cleaned observations were then assimilated into a high-resolution numerical weather model, allowing the model to update its description of the atmosphere as the storm evolved.

Data assimilation is a central technique in modern forecasting. Weather models calculate how temperature, pressure, moisture, wind, and other variables change according to physical equations. However, even the most advanced model begins with an imperfect representation of the real atmosphere. Data assimilation combines the model’s previous forecast with observations, weighting each according to its estimated reliability. In this study, smartphone pressure readings supplied additional information about the near-surface pressure field. That information helped the model adjust the storm’s structure and evolution, producing a more realistic simulation of the event.

The improvement was significant. Assimilating the smartphone observations increased hail forecast skill by approximately 14 to 17 percent and generated a better representation of where hail occurred and how the storm developed. In the Beijing case, the smartphone-derived measurements performed better overall than observations from traditional weather stations. Their advantage did not come from greater precision at any single location. Instead, it came from density: smartphones were distributed throughout populated areas, creating a finely spaced network capable of capturing pressure gradients and local changes that a sparse station network might overlook. In rapidly developing convection, such small-scale information can influence forecasts of storm intensity and location.

“Surface pressure contains useful information about the development and movement of convective storms, but traditional station networks cannot always observe these features in sufficient detail,” said Rumeng Li, the study’s corresponding author. The value of smartphones, Li explained, lies in their existing presence across cities. Establishing new meteorological stations requires land, equipment, maintenance, communications infrastructure, and long-term funding. By contrast, a smartphone-based observing system could potentially expand as more people use participating applications. If the data are collected responsibly and processed with rigorous quality control, the same devices that receive weather alerts could also help generate the observations behind them.

The approach nevertheless has important limitations. Smartphone measurements are not distributed evenly across the landscape. They are concentrated in places where people live, work, and travel, while forests, mountains, farmland, and sparsely populated areas may have few contributing devices. That imbalance was especially important in the Beijing storm because the system began developing over mountainous terrain, where smartphone observations were relatively scarce. The data improved the forecast after the storm moved into the city, but they could not fully correct errors in the earlier stages of storm formation. Smartphone observations also require careful handling of privacy, consent, location uncertainty, sensor calibration, and quality control before they can support operational warnings.

The researchers describe the Beijing analysis as an initial demonstration rather than proof that smartphones can replace conventional weather infrastructure. Future studies will need to test the method across many storms, climates, cities, and population distributions. Researchers will also need to determine how pressure data can be combined with radar, satellites, lightning networks, weather stations, and other sources of information. Even with those challenges, the concept offers an unusual route toward more localized forecasting. A phone network could provide high-frequency observations in places where conventional instruments are too expensive or too widely spaced, while also delivering warnings directly to the people most at risk. “Smartphones could help fill part of that gap by contributing pressure observations to forecast models,” said Qinghong Zhang, the project leader. The long-term vision is a cooperative system in which personal devices, meteorological stations, and radar work together to improve short-term predictions of hail, damaging winds, torrential rain, and other rapidly developing hazards.

Subject of Research: Smartphone-based atmospheric pressure observations for improving severe-weather and hailstorm forecasts.

Article Title: The Impact of Assimilating Dense Smartphone Pressure Observations on a Hailstorm Simulation

News Publication Date: 25-Jul-2026

Web References: https://doi.org/10.1007/s00376-026-5647-y

References: Advances in Atmospheric Sciences, DOI: 10.1007/s00376-026-5647-y

Image Credits: Rumeng Li

Keywords: Smartphones, weather forecasting, atmospheric pressure, hailstorms, severe weather, numerical weather prediction, data assimilation, machine learning, meteorology, early warning systems