How much more sun is there at 3,000 m?

Anyone who has been sunburnt on a ski slope knows the sun is stronger at altitude. Our solar irradiance endpoint takes an elevation, so we measured how much of that gain comes from the height and how much from the air.

Height by itself is the small part. Most of the gain comes from drier, cleaner air, and the endpoint has parameters for both.

What the model does with elevation

Our clear-sky figures come from the Bird model, published in 1981 by Richard Bird and Roland Hulstrom at what is now the US National Renewable Energy Laboratory.[1] It estimates how much sunlight reaches the ground under a cloudless sky by working through what removes light on the way down: scattering by air molecules, absorption by ozone and water vapour, and scattering and absorption by aerosols such as dust and haze.

Elevation enters through air pressure. At 3,000 m the pressure is about 710 hPa, against 1,013 hPa at sea level, so 30% of the air is below you. The model scales the optical air mass[2] by that ratio, which reduces the scattering by air molecules.

It leaves the water vapour and the aerosols alone unless you say otherwise. The defaults are 1.5 cm of precipitable water and an aerosol optical depth of 0.084, which describe a clean, fairly dry day near sea level.

The numbers

We computed global horizontal irradiance (GHI) for the sun 60° above the horizon on the March equinox, and changed one thing at a time.

Case GHI Change Direct beam (DNI)
Sea level, defaults 918 W/m² 918 W/m²
3,000 m, thinner air only 931 W/m² +1.4% 944 W/m²
3,000 m, and drier air 967 W/m² +5.3% 980 W/m²
3,000 m, drier and cleaner 994 W/m² +8.3% 1,081 W/m²
Sea level918 W/m²3,000 m, thinner air only931 W/m²3,000 m, and drier967 W/m²3,000 m, drier and cleaner994 W/m²Sea level, hazy890 W/m²Sea level, humid891 W/m²
Clear-sky global horizontal irradiance with the sun 60° above the horizon, from the Bird model. Height alone adds 1.4%, and drier, cleaner air adds the rest.

Height alone is worth 1.4%. Air molecules were never removing much light from a high sun, so taking 30% of them away does little.

The gain appears when you also account for what the air holds. Most water vapour and most aerosol sit in the lowest couple of kilometres, so a real site at 3,000 m is above most of both. For the last two rows we assumed 0.33 cm of water and an aerosol optical depth of 0.011. Those two values are our illustration of a mountain atmosphere. Use measured ones for your own site.

Field measurements have shown that solar ultraviolet gets stronger with altitude.[3] The model shows a gain with height too, once it is given the mountain's air.

It matters more for the direct beam, and when the sun is low

The direct beam responds far more than the total. In the last row DNI is up 18% while GHI is up 8%. Clean air scatters less light out of the beam, which raises DNI, and less scattered light means less arriving from the rest of the sky, which lowers the diffuse part. In GHI the two partly cancel.

If you are sizing a concentrating collector or a tracking panel, which use the direct beam, altitude and air quality matter a good deal more than the GHI figures suggest.

A low sun raises all of these. Its light crosses much more air, so removing some has a bigger effect. With the sun 30° above the horizon, the same four cases give +1.8% for height alone and +10.8% with drier and cleaner air, and DNI rises by 28%.

Haze costs more than height gains

For comparison, we went back to sea level and made the air worse.

Case GHI Change DNI
Sea level, defaults 918 W/m² 918 W/m²
Sea level, hazy (aerosol 0.20) 890 W/m² −3.1% 809 W/m²
Sea level, humid (4 cm of water) 891 W/m² −2.9% 891 W/m²

A hazy day at sea level takes away twice what 3,000 m of height gives back. If you care about accuracy, a good aerosol and water vapour figure for your site is worth more than a precise elevation.

What we take from this

Pass your elevation, because it is free and it is correct as far as it goes. Then, if your site is high, dry or dusty, pass the atmosphere too. The irradiance endpoint accepts water_cm and broadband_aod for that (the docs say which plans can set them). Left out, they default to the sea-level values above.

curl -H "Authorization: Bearer $THIRDTRAIL_KEY" \
  "https://thirdtrail.life/v1/solar/?lat=46.02&lon=7.75&elevation=3000&include=irradiance&water_cm=0.33&broadband_aod=0.011"

Irradiance needs a paid plan. Solar position is on the free tier.

Sources

  1. Bird, R. E., and Hulstrom, R. L. (1981). "Simplified clear sky model for direct and diffuse insolation on horizontal surfaces." Solar Energy Research Institute, SERI/TR-642-761. https://www.osti.gov/biblio/6510849
  2. Kasten, F., and Young, A. T. (1989). "Revised optical air mass tables and approximation formula." Applied Optics, 28(22), 4735. https://doi.org/10.1364/AO.28.004735
  3. Blumthaler, M., Ambach, W., and Ellinger, R. (1997). "Increase in solar UV radiation with altitude." Journal of Photochemistry and Photobiology B: Biology, 39(2), 130-134. https://doi.org/10.1016/S1011-1344(96)00018-8

Method: Bird clear-sky model with Kasten-Young air mass, pressure from the standard atmosphere, ozone 0.3 atm-cm, ground albedo 0.2, day 80 of the year. Clear-sky figures exclude cloud. Informational use only.

← All posts