Method and validation

How LogAloft works out the cosmic radiation dose of a flight, what goes into it, and how the numbers compare with measurements made on board aircraft. Last updated October 3, 2026.

• In short
• What is measured
• The model
• The flight path
• The Sun and the magnetic field
• Solar storms
• How it compares
• What it is not
• Sources

What is measured

LogAloft gives effective dose in microsieverts (µSv; 1,000 µSv = 1 mSv), with the weighting of ICRP Publication 103, the current international standard used by the EU directive on radiation protection and by the FAA. Effective dose is the number that adds up across sources: a flight, a chest X-ray and a CT scan are all expressed in it.

In Settings you can show two other quantities from the same calculation:

• ICRP 60 effective dose, the 1990 weighting some airlines still report. It comes out about 26–33% higher at cruise altitudes, mostly because it counts neutrons more. If your airline's figures run about 30% above ours, it is probably this.
• H*(10), ambient dose equivalent, what onboard dosimeters are calibrated to read: about 4–17% above ICRP 103 at cruise.

Both come from CARI-7A run with its other tallies at 1,488 points (31 places, 12 altitudes, 4 levels of solar activity), kept as ratio tables on the same grid as the doses. Your record is always kept in ICRP 103; the other two are shown from it.

The model

LogAloft uses CARI-7A (version 4.2.0), developed by the FAA's Civil Aerospace Medical Institute: galactic cosmic rays arriving at Earth, modulated by the Sun, sorted by Earth's magnetic field, and followed through the atmosphere with particle-shower tables calculated with MCNPX. Spain's Nuclear Safety Council recommends it for aircrew, and it is the reference in the FAA's guidance (AC 120-61B).

• CARI-7A was run ahead of time on a grid of 3,168 points: 24 levels of geomagnetic cutoff, 11 levels of solar activity, and 12 altitudes from 10,000 to 45,000 ft. A flight's dose is the dose rate read from that grid every minute along its path, added up.
• Below 10,000 ft, a curve from CARI-7A runs at low altitude, at polar and equatorial latitudes.
• Checked against CARI-7A run directly at 1,488 points: the grid agrees on average and within 1.5% everywhere.
• No shielding by the aircraft is applied, as airlines, regulators and the FAA do. The aircraft's structure and contents lower the dose a little: about 5% on the flight deck of an A340 model (Battistoni et al., 2005) and 12–16% on average in the cabin of a 777 model (Wang et al., 2023), so leaving it out errs on the high side. You can apply it in Settings.

The flight path

Dose depends most on how long you are at cruise, how high, and how far from the equator. LogAloft uses, in order of preference:

1. The track flown: positions and altitudes recorded by flight-tracking networks, fetched about 30 minutes after landing (Pro), or a GPS or avionics file you add. Gaps over the oceans are filled with the levels you enter or the published oceanic tracks.
2. The route for that day, for a flight not yet flown or with no track: the airways and oceanic tracks a dispatcher would plan for that day's winds, the filed cruise level when the flight number has one, and step climbs by aircraft type. Away from the ocean tracks, the path the same airline has flown between those airports in recent weeks, when we have its tracks. Around the airspace closed to that airline that day: the European regulator's conflict-zone bulletins (EASA) for European airlines, the FAA's prohibitions for US ones, by date (the list in use; FIR outlines from the VATSIM VAT-Spy Data Project, CC BY-SA 4.0).
3. Your own route and levels, when you type them in.

On 20 long-haul flights flown in September 2026, the planned path gave a dose within 3% of the dose along the real track on average; predicting the times and levels as well, within 9%. Across 42 flights of all lengths, the estimate before departure was off by 3 µSv on average. Each flight shows which of these it used.

The Sun and Earth's magnetic field

• Solar activity. The Sun's wind pushes galactic cosmic rays away; when it is active, doses at altitude fall, by up to about a third between solar minimum and maximum. LogAloft uses the FAA's monthly solar modulation values (heliocentric potential), and for months the FAA hasn't published yet, a nowcast from neutron monitors.
• Magnetic shielding. Earth's field turns away lower-energy particles, less so near the poles. LogAloft uses CARI-7A's maps of vertical cutoff rigidity for 1965 to 2010, carried forward to the flight's date; the field changes slowly.

Solar storms

A few times per solar cycle, the Sun sends out particles energetic enough to reach flight altitudes (a solar particle event, or a ground level enhancement). These add to the dose for hours, mostly on polar and high-latitude routes.

• The flight lookup shows today's space weather, and on a ground level enhancement adds an estimate from polar neutron monitors (the SiGLE method, calibrated on Concorde measurements).
• The logbook keeps galactic cosmic radiation as the record. In Settings you can also add solar radiation storms: LogAloft reads NASA's NAIRAS model every hour, keeps the hours when a storm lifts the dose (25% or more over the quiet level in the polar regions at 20 km, with NAIRAS's own flag when available), and adds the excess along each flight that flew through one. It goes back to the ground level enhancement of November 11, 2025. Storm doses are much less certain than the rest: models differed by 30–50% on past events, so LogAloft shows them separately on each flight.
• Geomagnetic storms (the aurora kind) weaken the magnetic shielding for a day or two; at FL350 they raise the dose rate by about 46–69% near the equator, depending on solar activity (about 59% under July 2026 conditions), by about 1% over North America to about 30% over East Asia at 45°N, and by under 10% from 55°N (next to nothing over North America). The flight lookup applies them, and so does the logbook when the storm setting is on (Kp 5 and up from GFZ Potsdam, the cutoff lowered as CARI-7A does it).

How it compares

The most careful published check of aviation dose models (Meier et al., 2018) used reference measurements of H*(10) made on board at known positions, altitudes and dates. LogAloft's numbers at those six points:

Dose rates in µSv per hour, H*(10). Average difference 5%, largest 10%; CARI-7A run directly averaged 6% in the paper.

• Between models. In the EURADOS comparison of eleven codes, most results fell within about 10% of the median and nearly all within ±30%; EURADOS asks any code to agree with measurements within 30%. Spain's regulator accepts results between 33% below and 50% above the real value.
• Against an airline's own report. A European airline's crew dose report for six long-haul flights was reproduced within about 5% with CARI-7A by using ICRP 60 weighting and cruise near FL410, which is what the ICRP 60 setting is for.
• On routes to South America. Iberia and Spain's CIEMAT measured 63 flights in 2001–2004 with a tissue-equivalent proportional counter (TEPC), 29 of them to South America, including Madrid–Bogotá and back (Sáez Vergara et al., 2004 and 2009). Run with the same inputs the paper gave the codes (date, flight time, mean altitude, great circle), LogAloft came out 6% below the measurements on average (±8%), every flight within the ±30% Spain requires, and in line with EPCARD, Iberia's own code (7% below). Madrid–Bogotá: 24.3 µSv against 24.4 measured.
• Against the ICRU reference values. ICRU Report 84 (2010) fitted about 20,000 onboard measurements from 1992–2006 and publishes reference dose rates for FL310, FL350 and FL390, cutoffs from 0 to 17 GV, in January 1998, 2000 and 2002. At those 162 points LogAloft is 4% below on average (from 20% below to 15% above) in H*(10), and 5% below in effective dose using the report's own conversion. Every point is within the ±30% the ICRU recommends for checking a dose calculation. By latitude: 5% above at high latitudes, 3% below at mid latitudes and 9% below near the equator, where LogAloft runs a little low.
• Newer models. Some newer codes (for example ACORDE) report higher doses on long flights. Their comparisons so far are against CARI-7A, not against measurements on board; LogAloft will add any model that is publicly available and checked against measurements.

What it is not

• Not a dosimeter. Nothing is measured on your flight; it is a model calculation, as airlines and authorities use.
• Not an official dose record. Where the law requires one (in the EU, for crew likely to exceed 1 mSv a year), it is your airline's. LogAloft gives you your own record to keep and compare.
• Not medical advice. What a dose means for your health is a question for your doctor; LogAloft's help center explains how doses compare with medical imaging and with legal limits.

Sources

1. FAA Civil Aerospace Medical Institute, CARI-7 and CARI-7A; Copeland, CARI-7A: development and validation, Radiation Protection Dosimetry (2017).
2. ICRP Publication 103 (2007) and ICRP Publication 60 (1990); on the difference between them, Meier and Matthiä (2019).
3. Sáez Vergara, Romero Gutiérrez, Rodríguez Jiménez and Domínguez-Mompell Román, In-flight measured and predicted ambient dose equivalent and latitude differences on effective dose estimates, Radiation Protection Dosimetry 110 (2004); Sáez Vergara and Domínguez-Mompell Román, The implementation of cosmic radiation monitoring in routine flight operation of IBERIA, Radiation Protection Dosimetry 136 (2009).
4. ICRU, Report 84: Reference Data for the Validation of Doses from Cosmic-Radiation Exposure of Aircraft Crew, Journal of the ICRU 10(2) (2010), Tables 6.1–6.4; the fit behind it: Wissmann, Reginatto and Möller, The ambient dose equivalent at flight altitudes: a fit to a large set of data using a Bayesian approach, Journal of Radiological Protection 30 (2010).
5. Meier et al., First steps toward the verification of models for the assessment of the radiation exposure at aviation altitudes, Space Weather (2018).
6. EURADOS Report 2012-03, Comparison of codes assessing radiation exposure of aircraft crew; EURADOS Report 2021-03 on solar particle events.
7. NASA, NAIRAS 3.0 (data from NASA's ISWA data tree); Mertens et al. (2013) on its validation.
8. Lantos and Fuller (2003), the SiGLE method for ground level enhancements, Radiation Protection Dosimetry.
9. Consejo de Seguridad Nuclear, exposure of aircrew to cosmic radiation; Council Directive 2013/59/Euratom, Article 35; FAA Advisory Circular 120-61B.

Data, credits and further sources

Dose engine: FAA CARI-7A (the model the FAA itself uses), ICRP-103 effective dose, galactic cosmic rays, no aircraft shielding. Magnetic cutoffs from CARI's own maps for 1965–2010, interpolated by date (the field drifts). Checked against CARI-7A's own flight integrator on 82 flights, worst case 0.8%. Solar modulation from the FAA’s monthly heliocentric potential table.

Solar activity: ground-level enhancements are estimated from polar neutron monitors (NMDB) with the SiGLE method (Lantos & Fuller 2003); Forbush decreases from the same data; Kp from GFZ Potsdam; protons from NOAA GOES.

Geomagnetic storms (Kp ≥ 5) weaken the magnetic shield: the cutoff rigidity is reduced the way CARI-7A does it (Al Anid's method), which at FL350 raises the dose rate by about 46–69% near the equator, depending on solar activity (about 59% under July 2026 conditions), by about 1–30% at 45°N depending on longitude (least over North America, about 22% over Europe), under 10% from 55°N, and by a few percent on North Atlantic routes (about 2.5% for New York–London); the flight's Kp comes from GFZ Potsdam.

For months the FAA has not published yet, the solar-cycle level is a nowcast from the Oulu neutron monitor (27-day mean count, fitted to the FAA table 2010–2026; about ±24 MV, ~1 % in dose), instead of carrying the last published month forward.

Limits: EU Council Directive 2013/59/Euratom (aircrew above 1 mSv a year are assessed individually; the unborn child ≤ 1 mSv for the remainder of the pregnancy; 6 mSv a year brings closer monitoring), FAA Advisory Circular 120-61B, paragraph 7c (0.5 mSv a month to the conceptus of a pregnant crewmember), ICRP Publication 132 (reference level 5–10 mSv a year for the most exposed), ICRP 103 (20 mSv a year occupational limit; about 1 mSv to the unborn child for the rest of a declared pregnancy).

Two numbers, two certainties. The galactic dose is good to about ±20% — the model is checked against dosimeters flown on hundreds of flights, the radiation field changes slowly, and the only inputs are where you were, how high, and the month. The solar-storm add-on is typically good to about ±50%, and for a single strong event it can approach a factor of two; that is the nature of the thing, not a shortcut here: a ground neutron monitor counts what reaches sea level, but the dose at FL350 depends on the storm's energy spectrum (a soft event barely shows on the ground yet dumps dose at altitude; a hard one is the reverse) and on its direction, which a count rate cannot tell. The method is calibrated on the few storms that were measured aboard aircraft — Concorde dosimeters in 1989 and 2000 — and the aircraft measurements of the same storm disagree with each other by that much. That is why the storm dose is shown on its own line and only added to the totals if you ask.

On-time records: US DOT (BTS). Fixes: national AIPs and the FAA CIFP. Airport elevations: OurAirports. Magnetic variation: the World Magnetic Model (NOAA and the British Geological Survey). Airline names: OpenFlights (ODbL). Weather: NASA GIBS (GOES, Himawari, IMERG), NOAA (nowCOAST global satellite mosaic; GOES-19 and GOES-18 rain rate), Iowa State Mesonet (NEXRAD), EUMETNET OPERA (radar, CC BY 4.0) and Deutscher Wetterdienst (radar, CC BY 4.0). NAT tracks: FAA, archived daily by this site. Winds: NOAA GFS; before 2021, ECMWF ERA5 (contains modified Copernicus Climate Change Service information).

Banner photo from Pexels (free to use under the Pexels license).

Runways in use, read from recent traffic where no ATIS is available: contains information from adsb.lol, made available under the Open Database License (ODbL) 1.0. Runway positions: OurAirports (public domain).

Cancer risk: BEIR VII (National Academies, 2006) lifetime attributable risk, linear no-threshold, age-adjusted; crew epidemiology from Cican & Silivestru, J. Space Safety Eng. 12 (2025) 615–630.

Offsets, directs and lateral deviations change the dose by well under 1%. Altitude and latitude are what matter.

Peer-reviewed sources. Copeland K. CARI-7A: development and validation. Radiat Prot Dosimetry 2017;175:419–431 · Bottollier-Depois JF et al. Comparison of codes assessing galactic cosmic radiation exposure of aircraft crew. Radiat Prot Dosimetry 2009;136:317–323 · Lantos P, Fuller N. History of the solar particle event radiation doses on-board aeroplanes using a semi-empirical model and Concorde measurements. Radiat Prot Dosimetry 2003;104:199–210 · Mertens CJ et al. Geomagnetic influence on aircraft radiation exposure during a solar energetic particle event in October 2003. Space Weather 2010;8:S03006 · Schennetten K et al. The impact of the Gannon storm of May 2024 on the radiation fields at aviation altitudes. Front Astron Space Sci 2024;11:1498910 · National Research Council. BEIR VII Phase 2, 2006 · ICRP Publication 132, Ann ICRP 2016;45(1) · Pukkala E et al. BMJ 2002;325:567 · Pukkala E et al. Int J Cancer 2012;131:2886–2897 · Hammer GP et al. Occup Environ Med 2014;71:313–322 · Blettner M et al. Int J Cancer 2003;106:946–952 · Sanlorenzo M et al. JAMA Dermatol 2015;151:51–58 · McNeely E et al. Environ Health 2018;17:49 · Schubauer-Berigan MK et al. Am J Ind Med 2015;58:252–266 · Tokumaru O et al. J Travel Med 2006;13:127–132 · Buja A et al. Toxicol Ind Health 2005;21:273–282 · Buja A et al. J Womens Health 2006;15:98–105 · Rafnsson V et al. Arch Ophthalmol 2005;123:1102–1105 · Cican F, Silivestru M. J Space Safety Eng 2025;12:615–630 · Patel V et al. JAMA Intern Med 2026. Data: FAA MV-DATES, NMDB, GFZ Potsdam, NOAA SWPC, US DOT Bureau of Transportation Statistics, FAA NOTAM system, NASA GIBS, Iowa Environmental Mesonet, national AIPs and FAA CIFP.

Questions about the method, or a measurement to compare with: contact support. See also About LogAloft and the help center.