UVA at altitude: mountains, hiking and flying

Everyone knows UV is stronger in the mountains. Rather fewer know that the familiar "10% more per 1000 metres" is an erythemal figure — it describes the sunburning band. UVA climbs with altitude too, but more gently, for reasons that say something useful about how the atmosphere works. And at altitude the gradient is rarely the biggest effect anyway.

The correction: the standard 8–10% per 1000 m applies to erythemal UV. For UVA specifically the gradient is more like 4–7% per 1000 m. This site's UVA Index uses +6% per kilometre, deliberately below the erythemal number.

Why altitude raises UV at all

Go up and there is simply less atmosphere between you and the sun. At 3000 m roughly 30% of the atmosphere's mass is already below you. That reduces four separate things at once:

Why UVA's gradient is gentler

The gradient differs by wavelength because the things you leave behind are not wavelength-neutral.

Ozone absorbs strongly in the UVB and is almost transparent in the UVA (the same asymmetry that flattens UVA's seasonal cycle). Losing ozone from your path is a large gain for UVB and a negligible one for UVA.

Rayleigh scattering scales roughly as λ−4 — it is dramatically stronger at short wavelengths. That factor alone makes molecular scattering about 2.5 times stronger at 300 nm than at 380 nm. Climbing above some of the atmosphere therefore recovers much more UVB than UVA.

What is left is aerosol, which is far more wavelength-neutral across the UV, and it is the mechanism that UVA benefits from most. That has a practical implication: in a hazy or polluted region the UVA altitude gradient is steeper than in clean air, because there is more haze to escape. A climb from a smoggy valley floor is a bigger UVA jump than the same climb in the Arctic.

What the numbers look like

Using the model's +6%/km, and adding the surface and air effects that travel with altitude in practice:

SettingElevationAltitude factorWith snow albedo
Sea level0 m×1.00
Alpine valley town1000 m×1.06×1.17
Ski resort base2000 m×1.12×1.23
Ski resort summit3000 m×1.18×1.30
High trekking pass5000 m×1.30×1.43

The altitude term on its own is real but not dramatic. The reason mountains are a high-UVA environment is that altitude never arrives alone. At a ski summit you have +18% from elevation, +10% from snow reflecting from below, and a clean, dry, low-aerosol atmosphere that costs you less than sea-level haze would. The compound effect is what makes a clear February day at 3000 m rival a summer day at sea level — despite a solar elevation that ought to make it four times weaker.

Why the sensory cues fail completely here: altitude raises UV while lowering temperature. Every instinct you have for judging sun intensity is calibrated on warmth, and at 3000 m in winter that calibration is not just wrong, it is inverted.

Aircraft: cockpits, cabins and the pilot data

At a cruise altitude of 10–12 km you are above roughly three quarters of the atmosphere, and UV intensity is several times its sea-level value. What reaches the people inside depends entirely on the glazing.

The epidemiology is what pushed this into the literature. A meta-analysis of aircrew studies found roughly twice the melanoma incidence in both pilots and cabin crew compared with the general population, and higher melanoma mortality in pilots. The finding is consistent across studies and hard to dismiss.

It is worth being careful about what it proves. Aircrew differ from the general population in more than their working altitude: they have high incomes, frequent access to sunny destinations, disrupted circadian rhythms, cosmic-radiation exposure and — importantly — better-than-average medical surveillance, which raises detected incidence on its own. Most authors regard occupational UV as a plausible contributor rather than a demonstrated sole cause. The cockpit UVA measurements make the mechanism credible; they do not by themselves close the case.

How the UVA Index handles elevation

The calculator does not ask you for your altitude. It looks up the terrain elevation for your coordinates from Open-Meteo and applies:

altitude factor = 1 + 0.06 × (elevation_m ÷ 1000)

A linear approximation is a simplification — the real relationship flattens somewhat at very high elevations — but it is well behaved across the range where people actually stand, and the coefficient is UVA-specific rather than borrowed from erythemal tables. You can see the factor itself in the calculator's Model breakdown, alongside the cloud, aerosol and albedo terms; the derivation for all of them is in how the UVA Index is calculated.

Two limits to keep in mind. The elevation used is the ground elevation at your coordinates, so the model has nothing sensible to say about being in an aircraft. And it does not know that mountain air is typically cleaner than the aerosol data suggests for the wider grid cell, so on a clear alpine day the estimate is more likely to run low than high.

Practical takeaways

The UVA Index uses your real terrain elevation →