Solar Air Mass: AM1.5 Explained With Formulas

Last updated: November 2026

Every solar module datasheet lists the same test conditions: 1,000 W/m² irradiance, a 25°C cell, and the AM1.5G spectrum. Those three letters hide one of the most useful ideas in PV design. Solar air mass is the relative length of the atmospheric path that sunlight travels before it hits the array. It explains why a panel in Denver at noon can beat its nameplate rating. It also explains why a panel in Helsinki in December cannot. And it shows why the 1,000 W/m² benchmark is a convention, not a promise.

In 2026, the gap between lab ratings and real rooftops still drives sizing errors. Designers who treat Standard Test Conditions as a site forecast overstate peak irradiance. They undersize inverters for high-altitude sites and overstate winter output at high latitudes. This guide gives the engineering answer first, then shows how to apply it in practice.

In this guide:

• What solar air mass means and why AM1.5G is the global standard
• How to calculate air mass from zenith angle, latitude, and altitude
• The difference between AM0, AM1, AM1.5G, and AM1.5D
• How air mass, aerosols, and pressure change the spectrum at your site
• Why AM1.5G alone is not enough for high-precision yield modeling
• How to check the air mass and spectral assumptions in a yield model

What Is Solar Air Mass?

Solar air mass is the ratio of the atmospheric path length that sunlight travels to the shortest possible vertical path at sea level. The shorthand is AM, and the number tells you how many atmospheres the beam has passed through.

At solar noon on the equator at equinox, the sun sits close to the zenith. The light passes straight down through one atmosphere. That condition is AM1. At a mid-latitude site with the sun 48.2° from the vertical, the path is 1.5 times as long. That condition is AM1.5. Near sunrise or sunset the path can exceed AM38, although most of the useful daily energy arrives at much lower air masses.

The atmosphere is not empty space. Air molecules, water vapor, aerosols, and ozone scatter and absorb photons. Blue light scatters strongly. Infrared light is absorbed by water vapor. The longer the path, the more energy is lost and the more the surviving spectrum shifts toward red and near-infrared. This matters because a PV cell does not weigh every photon equally. Silicon devices respond from roughly 300 nm to 1,200 nm. A spectrum with less blue content produces less current, even if the total wattage stays the same.

The concept is explained clearly by PV Education, which shows the geometry and the standard secant approximation.

The term air mass confuses newcomers because meteorologists use it differently. In weather forecasting, an air mass is a large body of air with similar temperature and humidity. In solar engineering, air mass is a pure geometric ratio. The two definitions are unrelated. When a datasheet says AM1.5G, it is talking about path length, not weather.

AM1.5G vs AM1.5D vs AM0: The Standard Spectra Explained

Module datasheets use AM1.5G because the solar industry needed a single, repeatable reference. The G stands for global, meaning the spectrum includes direct beam, diffuse sky radiation, and ground-reflected radiation on a tilted plane facing the equator. The D variant stands for direct and is used for concentrating systems that track the sun and reject diffuse light.

The integrated powers above are listed on PV Education’s standard spectra page. The AM1.5G reference is defined by ASTM G173-03 and published by NREL. The Sandia PV Modeling Collaborative explains how AM1.5G became the IEC 60904-3 reference spectrum. The integrated irradiance under the AM1.5G table is 1,000.37 W/m², which is why the 1,000 W/m² STC number is so familiar.

The standard tilts the plane to roughly the average latitude of the contiguous United States, which is a reasonable average for temperate zones. It is not the average for Singapore, Stockholm, or Santiago. That single choice is the source of much confusion. A module rated at AM1.5G will perform slightly differently under a tropical noon. The spectrum there is harder and the air mass is lower. It will also perform differently under a Nordic winter, where the air mass is thicker and the spectrum shifts red.

The AM1.5G standard has a long history. The original ASTM E-891 and E-892 tables from 1982 were combined and revised into ASTM G159, then replaced by ASTM G173-03 in 2003. The International Electrotechnical Commission adopted the same spectral basis in IEC 60904-3. Today, every module flash tester and every bankable energy model points back to this single reference. Real spectra still vary from site to site and hour to hour.

How to Calculate Solar Air Mass for Any Site

For most design work you never compute air mass manually. Your solar design software calculates it thousands of times per year behind the scenes. But the formula is short, and understanding it prevents mistakes.

First find the solar zenith angle θz. For a given latitude φ, solar declination δ, and hour angle ω:

cos θz = sin φ sin δ + cos φ cos δ cos ω

The hour angle is 0° at solar noon and changes by 15° per hour. The declination varies from +23.45° at the June solstice to -23.45° at the December solstice.

For zenith angles below about 70°, the simple secant approximation is accurate enough:

AM ≈ 1 / cos θz

At larger zenith angles the curvature of the Earth and atmospheric refraction matter. The Kasten and Young 1989 formula is widely used:

AM = 1 / [cos θz + 0.50572 × (96.07995 − θz)^−1.6364]

where θz is in degrees. A good walkthrough of these approximations is given by G2V Optics.

At altitude, correct for pressure:

AM_absolute = AM_relative × (P_site / 1013.25 hPa)

In practice, designers do not look up declination tables. Solar design software and libraries such as pvlib use the NREL Solar Position Algorithm. They compute θz for every hour of the year from latitude, longitude, altitude, and timezone. The air mass follows from that. If you are checking a result by hand, the simple secant formula is enough for zenith angles below 70°. For sunrise, sunset, or high-latitude winter, switch to Kasten and Young.

Worked example: Denver at solar noon, March equinox

Denver sits at 39.74°N and 1,609 m above sea level. On the equinox the solar declination is 0°, and at solar noon the hour angle is 0°. The zenith angle equals the latitude: 39.74°.

• cos(39.74°) = 0.769
• Relative AM = 1 / 0.769 = 1.30
• Typical Denver pressure at 1,609 m ≈ 834 hPa
• Absolute AM = 1.30 × (834 / 1013.25) = 1.07

A sea-level site at the same latitude would see AM1.30. Denver effectively sees AM1.07 because the atmosphere above it is thinner. That is why high-plains arrays can experience clear-sky irradiance above 1,000 W/m² at solar noon.

Worked example: Mumbai at solar noon, March equinox

Mumbai sits at 19.08°N and about 14 m above sea level. At solar noon on the equinox the zenith angle is approximately 19.08°.

• cos(19.08°) = 0.945
• Relative AM = 1 / 0.945 = 1.06
• Sea-level pressure correction is essentially 1.0
• Absolute AM ≈ 1.06

Mumbai is close to AM1 at solar noon. The beam passes through little atmosphere, so the spectrum is harder and the clear-sky irradiance is high. The gap between Mumbai and Denver is small in air mass but still meaningful for spectral correction and inverter clipping studies.

Why Air Mass Matters in PV System Design

Air mass changes two things at once: total irradiance and spectral distribution. Both affect yield.

Total irradiance falls because the atmosphere scatters and absorbs photons. The AM1.5G spectrum integrates to about 1,000 W/m², roughly 27% below the AM0 reference of about 1,366 W/m². On a clear day with lower air mass, the direct-normal irradiance can approach 1,000 W/m² or slightly more. On a high-air-mass winter afternoon it may fall below 400 W/m² even with a clear sky.

Spectral distribution matters just as much. A silicon cell converts photons to current only when the photon energy exceeds the bandgap. Higher-energy photons are absorbed near the surface; lower-energy photons pass deeper. At high air mass the blue end of the spectrum is depleted. The short-circuit current drops even if a pyranometer reports the same total irradiance.

The size of the effect depends on module technology. Crystalline silicon is generally less sensitive to spectral shifts than thin-film or multi-junction devices, but it is not immune. The correction is small on a clear mid-latitude day and grows at high air mass, so it matters most for winter, high-latitude and bankable energy models.

Air mass also interacts with temperature. A low-air-mass site tends to have high irradiance, which raises cell temperature. The temperature coefficient then reduces voltage and power. A high-altitude site can have low air mass and low ambient temperature at the same time. The module runs cooler and produces more power than the AM1.5G rating suggests. You cannot look at air mass in isolation.

A common misconception to avoid

AM1.5G is not the average noon condition worldwide. It is a standardized test spectrum chosen to represent a reasonable mid-latitude average. At March-equinox solar noon, Mumbai sits near AM1.06 (worked example below). At December solar noon in Berlin the air mass is above 4. Treating AM1.5G as a universal site condition is one reason STC-only spreadsheets mislead clients.

Our view on spectral corrections

For commercial and utility-scale projects, ask which spectral and incidence-angle losses the yield model applies and where the assumption is shown in the report. A model that starts from STC power and a generic sun-hour map gives you no way to audit that step.

Altitude, Aerosols, and Real-World Air Mass Corrections

Clean-air geometry is only the starting point. The real atmosphere adds aerosols, humidity, and pollution.

Altitude lowers barometric pressure. Less atmosphere means less scattering and a harder spectrum. The pressure-corrected example above showed Denver at AM1.07 versus a sea-level AM1.30 for the same zenith angle. In the Andes or the Himalayas the effect is even larger. At high elevation, clear-sky irradiance can sit above sea-level expectations, and cold, clear hours can push output past module STC power.

Aerosols do the opposite. Dust, smoke, haze, and urban pollution increase effective optical depth. The actual spectrum becomes redder and dimmer than the clean-air AM model alone predicts. The Middle East, North India, and parts of sub-Saharan Africa see heavy seasonal dust. Heavy aerosol loading adds optical depth on top of the geometric air mass. Satellite aerosol optical depth products, such as those embedded in high-quality weather files, capture this better than a textbook formula.

Clouds dominate in many climates. A thick cloud layer does not have a single air mass; it scatters most of the direct beam into diffuse radiation. That is why air mass is most useful under clear-sky or partly cloudy conditions. Yield models still rely on measured or satellite-derived irradiance, not pure geometry.

Water vapor absorbs infrared. High humidity removes photons in the 900–1,200 nm range, where silicon still has response. This matters in tropical and coastal climates. Precipitable water is another input that clean geometry ignores but good yield models include.

The practical rule is simple: use a site-specific Typical Meteorological Year or satellite-derived weather file. Do not rely on AM1.5G as a proxy for local noon.

Air Mass, Angle of Incidence, and Module Rating Gaps

Air mass and angle of incidence are related but distinct. Air mass describes the sun’s path through the atmosphere. Angle of incidence describes how tilted the module surface is relative to the incoming beam. A tracker can keep incidence angle low while air mass still rises in the morning and evening. A fixed-tilt array at high latitude may face a high air mass and a high incidence angle at the same time in winter.

This distinction is why nameplate kWp is a benchmark, not an operating forecast. Take Berlin as an illustration. At solar noon in June the zenith angle is about 29°, so air mass is near 1.15. At solar noon in December it is about 76°, so air mass is above 4 and the sun sits low. Clear-sky irradiance and module output at that hour are a small fraction of the June value, and angle-of-incidence losses rise.

Designers who sell systems using STC peak power as a daily expectation set clients up for disappointment. The right metric is annual specific yield, expressed as kWh per kWp, computed with hourly air mass, temperature, incidence angle, and soiling models.

Where Air Mass Fits in a Yield Workflow

Hand correction is fine for a classroom exercise. A real proposal has too many hours and variables for it. Air mass is one input inside the yield model, alongside irradiance, temperature, shading and incidence angle.

SurgePV supports energy-yield modeling, shading analysis and financial modeling in one workflow. As with any yield tool, results depend on source data, assumptions, equipment models and configuration, and need engineering review. If you want to see which loss assumptions a model uses for your site, look at the solar design software workflow in a demo. Shadow analysis covers the obstruction and shading side, and the generation and financial tool turns the yield into cash flows.

A practical checklist for reviewing any yield model:

1. Confirm the weather file is site-specific (a Typical Meteorological Year or satellite-derived data).
2. Check that site elevation or pressure is an input, not an afterthought.
3. Find the stated spectral and angle-of-incidence loss assumptions.
4. Compare the annual specific yield (kWh per kWp) with a second method.
5. Record the assumptions in the proposal so the customer sees them.

Frequently Asked Questions

What is solar air mass?

Solar air mass is the ratio of the atmospheric path length that sunlight travels to the shortest possible vertical path at sea level. AM1 means the sun is directly overhead. AM1.5 means the light passes through 1.5 times as much atmosphere, which corresponds to a solar zenith angle of about 48.2°.

Why is AM1.5G used for solar panels?

AM1.5G is the global reference spectrum for rating flat-plate PV modules. It represents hemispherical irradiance, direct plus diffuse, on a tilted surface facing the equator. The integrated irradiance under AM1.5G is 1,000 W/m², which matches the Standard Test Conditions used on every datasheet.

What is the difference between AM1.5G and AM1.5D?

AM1.5G is global hemispherical radiation. It includes direct beam, sky diffuse, and ground-reflected light. AM1.5D is direct normal radiation only, plus a small circumsolar component. Flat-plate modules use AM1.5G; concentrating PV and solar trackers that focus direct light use AM1.5D.

How do you calculate solar air mass?

For zenith angles below 70°, air mass is approximately 1 divided by the cosine of the zenith angle. For all angles up to 90°, the Kasten and Young 1989 formula is more accurate. At altitude, multiply the relative air mass by the ratio of local barometric pressure to 1,013.25 hPa.

Does altitude change air mass?

Yes. Higher altitude means lower atmospheric pressure and less scattering. A site at about 1,600 m, like Denver, has an absolute air mass roughly 18% lower than a sea-level site at the same zenith angle, because pressure there is about 834 hPa against 1,013.25 hPa. That is one reason high-altitude arrays can exceed STC output at solar noon.

Does air mass affect solar panel efficiency?

Air mass changes both the intensity and the spectral distribution of light. Higher air mass reduces total irradiance and removes more blue light. Because silicon cells respond best to roughly 300–1,200 nm, the spectral shift changes short-circuit current even when total irradiance is the same.

How do designers use air mass in practice?

Designers rarely calculate air mass by hand. Solar simulation tools compute hourly air mass from solar position, site pressure, and weather data. It then applies angle-of-incidence, temperature and other loss models to predict actual yield instead of STC output. Check the loss assumptions in your tool, because spectral treatment varies.

What to Do Next

• Audit your next three proposals. Check whether yield estimates are based on STC assumptions or on hourly air-mass-adjusted models.
• Add site elevation and local aerosol data to your design inputs. A sea-level AM formula will mislead you in Denver, Mexico City, or Kathmandu.
• Run the same project through a second tool and compare the production forecasts. Differences usually trace back to spectral, incidence-angle or elevation assumptions.

Want to compare your current numbers against a SurgePV model? Book a demo and bring one live project.