In October 2025, a JetBlue Airways A320 flying from the United States to Mexico suddenly lost altitude data and made an emergency landing. Fifteen passengers were injured. The cause was traced to high-energy particles arriving from the sun, which corrupted data in the flight control computer responsible for calculating altitude. In response, Airbus issued an urgent request on November 28 for a software fix to be applied to roughly 6,000 aircraft in the A320 family. Airlines around the world scrambled to comply over the weekend.
This incident demonstrated that space weather can have a direct impact on aviation safety and operations. That said, radiation exposure from cosmic rays is not a new issue for the aviation industry. Airliners cruise at altitudes of 8–14 km, where atmospheric shielding is far thinner than at ground level. Galactic cosmic rays (GCR) continuously collide with atomic nuclei in the upper atmosphere, generating secondary particle showers such as neutrons and muons. The effective dose a passenger receives on a one-way flight from London to New York is about 0.08 mSv — roughly equivalent to a single chest X-ray.
The real concern, however, is not this "everyday" exposure but the sudden radiation spikes triggered by the sun.
Why Current Alerts Keep Missing the Mark
For space weather phenomena, NOAA's Space Weather Scale (S-scale) has effectively served as the international standard for alerts. The S-scale classifies events from S1 to S5 based on the flux of protons with energies above 10 MeV. But what actually drives up radiation levels at aviation altitudes is a phenomenon called a "Ground Level Enhancement" (GLE) — secondary particle showers produced by nuclear reactions in the atmosphere from high-energy particles exceeding 500 MeV.
Professor Clive Dyer of the University of Surrey points out this mismatch clearly. Current alerts are based on low-energy proton events, resulting in "at least ten false alerts for every true alert." GLE 77 on November 11, 2025, was classified as S2 on the S-scale, but in terms of aviation radiation it corresponded to AR2/3. Conversely, some events that reached S4 registered as AR0 (negligible) from an aviation radiation standpoint. In other words, the current scale has been structured in a way that, for aviation purposes, it fails to sound the alarm when it should — and sounds it when it doesn't need to.
Furthermore, the International Commission on Radiological Protection (ICRP) treats aircrew exposure as occupational exposure and recommends an annual limit of 20 mSv (averaged over five years), but the possibility that an Extreme Solar Particle Event (ESPE) could exceed this limit in a single flight has not been built into operational rules.
The "Reference Event" of 1956 Reveals the Ceiling
The foundation of this study's quantitative assessment is GLE 5, which occurred on February 23, 1956. This is known as the largest solar proton event on record since the era of neutron monitor observations began. The neutron monitor at Leeds, UK, recorded a 4,760% increase over the normal count rate from galactic cosmic rays.
Dyer's team fed the spectral model of this event into the MAIRE+ model to calculate dose rates at flight altitude and single-event upset (SEU) rates in electronic equipment. The results show that, at flight altitude over the UK, the dose rate reached about 5 mSv/hour — roughly 1,000 times (three orders of magnitude) the normal value. If this event lasted 1.6 hours, passengers would have received about 8 mSv in a single flight — equivalent to the exposure from about 100 ordinary flights (roughly a year's worth) all at once.
As for SEU rates, extreme ESPEs can reach 6,000 to 20,000 times the normal background level. For even larger events estimated by past reconstruction models, this could reach up to 100,000 times. An SEU occurs when a high-energy particle flips a bit in a memory cell; if these accumulate, they can cause faults in flight control and communication systems.
| Metric | Normal (GCR background) | GLE 77 (November 2025) | GLE 5 (February 1956, estimated) |
|---|---|---|---|
| Dose rate at flight altitude | ~7–8 μSv/hour | >30 μSv/hour (~4x) | ~5 mSv/hour (~1000x) |
| SEU rate (polar regions) | Background level | ~60 errors/hour/Gbyte | 1,000–100,000x background |
| Passenger exposure per single flight | ~0.08 mSv (transatlantic) | ~4x normal | Equivalent to a year's worth of flight exposure |
| S-scale classification | None | S2 | Not applicable (outside S-scale) |
What Balloons and Neutron Monitors Actually Measured
On November 11, 2025, an X5-class solar flare occurred, triggering GLE 77 — the largest GLE in roughly 25 years. The team at the University of Surrey and the UK Met Office did not let this opportunity pass. They simultaneously launched weather balloons from three sites — Shetland, Cornwall, and De Bilt in the Netherlands — carrying Surrey's radiation sensor "SAIRA," and obtained real-time radiation profiles from ground level up to over 30 km (100,000 feet) in altitude.
This is the first time radiation data has been collected across such a broad range of altitudes during a space weather event. Krista Hammond, the Met Office's space weather manager, said, "It's the first time we've gathered radiation data across such a wide range of altitudes, from the ground up through the atmosphere, and it's a major step forward for space weather forecasting capability."
At 40,000 feet (about 12 km, an airliner cruising altitude), radiation levels briefly reached about 10 times normal. However, this event was only about 2% the magnitude of the 1956 event, and did not reach a level posing an immediate health risk.
The Structure of the Proposed AR Scale
The AR scale (Atmospheric Radiation scale) proposed by the research team consists of six levels, from AR0 to AR5. Its biggest difference from the S-scale is that it is based on quantities that are both measurable and directly, physically related to aviation radiation. Specifically, it uses the increase in count rate from ground-level neutron monitors (GLNM) and flux data from proton channels above 500 MeV, such as those from GOES satellites.
Each level comes with a corresponding recommended action. At AR2 (Moderate), enhanced monitoring is advised; at AR3 (Strong), reconsideration of polar routes is recommended; at AR4 (Severe), avoiding flight operations is strongly recommended; and at AR5 (Extreme), halting operations is mandatory. AR5 corresponds to an event on the scale of the 1956 event, estimated to occur once every 50–70 years. AR4 represents an event about half that magnitude, occurring roughly once every 30 years.
Professor Dyer states that "recommended actions are purely suggestions and need detailed discussion with all stakeholders," including pilots, indicating that operational implementation remains a task for the future.
Data from 5 Million Flights Reveals a Correlation Between Space Weather and Cancellations
Evidence that space weather affects aviation operations extends beyond radiation exposure and electronics failures. Another study, published in Scientific Reports in 2026, analyzed departure records from approximately 5 million flights at five hub airports in China between 2015 and 2019, and reported that the flight cancellation rate during space weather events rose from 2.30% during quiet periods to 4.53% — an increase of about 97%. When limited to solar proton events (SPE) specifically, the increase rate reached 137%. This is the first systematic evidence showing that space weather affects operations at mid-latitude airports as well, not just on polar routes.
Questions That Remain
While the AR scale is conceptually clear, unresolved challenges lie on the path to operational implementation. First, there is the issue of the GLE's time profile and anisotropy. The peak dose rate of the 1956 event is estimated to have lasted 1.6 hours over the UK, but neutron monitor data from Ottawa suggests that the same cumulative dose accumulated over more than 4 hours — indicating that the event's spatial structure can change how the impact is distributed over time.
Second, although the MAIRE+ model employs a physics-based data assimilation approach, no actual measured data exists for an event on the scale of the 1956 event. Validation of the model must rely on comparison with smaller-scale events such as GLE 77 in November 2025.
Third, ICAO held a workshop on radiation exposure in the North Atlantic region in March 2026, where this AR scale was put on the table for discussion, but it has not yet been adopted as a regulation. Numerous non-technical issues remain unresolved, including the economic criteria airlines would use to decide on halting operations, pilot training requirements, and how such events would be treated for insurance purposes.
The sun is currently past the peak of Solar Cycle 25. When the next GLE 5-class event occurs, will the aviation industry have the AR scale in hand? That will determine whether this study's proposal remains merely a theoretical framework — or becomes a tool that actually protects the skies.