EN

Solar Incidence Angle Calculator (Tilt & Orientation)

A beam of parallel rays — like sunlight — only delivers its full power onto a surface that faces it perpendicularly. As soon as the surface tilts away from that direction, it intercepts fewer rays per unit of its own area: the fraction actually captured follows the cosine law, cos θ, where θ is the angle of incidence between the rays and the surface's normal.

This purely geometric relationship applies to any surface exposed to the sun: a photovoltaic panel, skin, a wall, a shop window, a Trombe wall. What changes afterwards is only how each one absorbs the energy it receives — which this page doesn't calculate.

This page calculates the geometric factor only. For the loss due to travelling through the atmosphere depending on the sun's elevation, see the Solar Atmospheric Attenuation page — the two factors multiply together to give the power received by a fixed surface.

Surface, location, date and time

The surface

0° = horizontal surface (laid flat), 90° = vertical surface (a wall).

Direction the surface faces. Due south is optimal in the northern hemisphere.

Location, date and time

Positive in the northern hemisphere, negative in the southern hemisphere.

Format MM-DD, e.g. "06-21" for the summer solstice. The year doesn't affect this calculation.

True solar time (12:00 = solar noon, sun at its highest) — not the time on a watch, which shifts with time zone and longitude.

β = 30°nθ = 30°87 %captured
Principle diagram (example values). The diagram redraws with the values entered after the calculation.

A star after a word leads to its glossary entry.

Lambert’s cosine law for a parallel beam

Sunlight at the ground, on a clear day, behaves locally as a bundle of parallel rays. A surface facing those rays square-on intercepts the most power per unit of its own area. Tilt the surface and the same bundle is spread over a larger patch. The factor is the cosine of the angle of incidence* θ, the angle between the incoming ray and the surface normal n.

Ireceived = Inormal · cos θ

That is Lambert’s cosine law* for a parallel beam (the same cosine appears in Lambert’s emission law, for a different reason). It is a dot product: the flux through a patch is n · S, and n · S = |S| cos θ. Geometrically it is the projected area of the patch as seen along the rays. At θ = 0°, cos θ = 1. At 60°, half the power. At 90°, grazing, nothing. Beyond 90° the front face is in its own shadow; this page reports zero direct capture.

The identity used here, after the Sun’s elevation α and azimuth are known, is

cos θ = sin α · cos β + cos α · sin β · cos(Az − γ)

β is the tilt from the ground, γ the azimuth the surface faces. The formula is geometry only. It does not know about glass, cells, or skin. Those come after, as absorption.

Why a fixed array is often tilted near the site latitude

Over a year the Sun’s mean direction, at a given latitude φ, sits roughly φ away from the vertical. A fixed panel whose normal points that way — tilt β ≈ φ, facing the equator — keeps the annual average of θ smaller than a flat roof or a wall would. At the equator, β ≈ 0°. Near the Arctic Circle, a steep tilt. Between those, latitude is the usual first guess.

It is a guess, not a theorem. Cloudier winters, a heating load that peaks when the Sun is low, or a tariff that pays more in the evening, all shift the economic optimum. Rule-of-thumb bands of φ ± 10°, or φ − 15° in summer and φ + 15° in winter for a seasonally adjustable rack, are engineering folklore with a grain of truth. A site with a dark December and a bright June may prefer a slightly shallower tilt than φ, to harvest the long days.

The solar panel output calculator turns that geometry into watt-hours, once efficiency and area are named. This page stops at cos θ.

South-facing versus east/west: how much annual yield really moves

In the northern hemisphere, due south puts solar noon at the centre of the day. An eastward azimuth moves the peak to the morning; westward, to the afternoon. Over a full year, a ±30° miss in azimuth often costs only about 5% to 10% of annual beam capture on an otherwise well-tilted plane, because the Sun’s path is wide. Winter, with a short arc, is less forgiving. Summer barely notices.

Roofs are not compasses. An existing 20° west-of-south pitch is usually kept. The cosine law is what says that decision is cheap. A north-facing pitch in the UK is another matter: θ stays large for the beam, and the surface lives on diffuse sky, which this page does not model.

East-west split arrays (two faces, one each way) trade a lower noon peak for a broader daily plateau. The incidence angle on each face is worse at noon and better at one end of the day. Grid-constrained sites sometimes prefer that shape to a single south face.

Single-axis and dual-axis tracking

A tracker’s job is to keep θ near 0°. A single-axis tracker, commonly a north-south horizontal axis, follows the daily east-west motion. A dual-axis tracker follows elevation as well as azimuth. Typical gains over a well-tilted fixed rack, in beam-rich climates, sit around +20% to +25% for a single axis and +30% to +40% for two axes. The extra is largest when a fixed surface would be seeing a large θ: morning, evening, winter.

At solar noon on a well-aimed fixed plane, cos θ is already close to 1. The tracker has little to add in that hour. It earns its keep in the shoulders of the day. Concentrating collectors, which cannot use diffuse light, need that tracking; a flat photovoltaic module does not, but benefits all the same.

Trackers add motors, wind loading, and land. The cosine law does not decide whether they pay. It only says what θ would have been without them.

Skin and UV: the same cosine

A square centimetre of skin is a surface with a normal. Direct UV irradiance on that patch is scaled by cos θ, just as on a panel. At a summer noon with the Sun near the zenith, θ is small on the top of the head and on level shoulders. Late afternoon, θ on those same patches climbs toward 90°, and the direct dose falls even if a published UV index (defined on a horizontal plane) looks moderate.

Vertical skin — a cheek, a shirt-front — has a different θ from a forehead. That is why the nose burns on a spring walk while the scalp, under hair, may not: different normals, different cosines, different shade. The cosine does not replace a UV index. It explains why the same index feels harsher at noon than at four o’clock on an open field.

Trombe walls and south glazing through the seasons

A Trombe wall* is a dark masonry wall behind glazing, used as a passive solar heater. In winter the Sun is low. On a vertical south window, θ is then relatively small around noon: rays arrive not far from perpendicular, cos θ is high, and the wall drinks heat. In summer the Sun is high. On that same vertical glass, θ is large, cos θ is smaller, and a bigger share of the beam is reflected or never aimed at the wall. Overhangs push the summer story further.

South-facing glazing in a temperate climate is therefore not a seasonal contradiction. The elevation of the Sun does part of the shading. Summer needs ventilation and sometimes an actual shade; the cosine is a help, not a complete design. The same geometry is why a horizontal skylight is a summer oven and a weak winter collector: the opposite of the wall.

Multiplying incidence by atmospheric attenuation

cos θ dilutes the beam across the surface. The beam itself is already weaker than it was above the atmosphere, because it has crossed air. The two factors multiply:

Isurface = DNI · cos θ

DNI, direct normal irradiance, is the leftover after atmospheric attenuation, on a plane facing the Sun. The solar atmospheric attenuation page isolates that path-length effect. This page isolates θ. A fixed roof needs both, plus a diffuse term this geometry does not compute.

Air mass grows as the Sun sinks, so a late-afternoon ray is both more attenuated and, on a fixed south face, often at a worse θ. The two penalties stack. A tracker removes most of the θ penalty and cannot remove the air mass.

True solar time, the equation of time, and longitude

The Sun’s hour angle ω is 15° per hour from solar noon. Clock noon is not solar noon. Two corrections sit between them: longitude relative to the time-zone meridian (4 minutes per degree), and the equation of time*, which wanders by up to about ±16 minutes through the year because the orbit is elliptical and the equator is tilted.

This form asks for true solar time. Feeding watch time without those corrections shifts ω, hence α and azimuth, hence θ. Four minutes is 1° of hour angle. In mid-morning or late afternoon, when cos θ is changing fast, that 1° is visible on the percentage captured. At solar noon on a due-south plane it is almost invisible.

The astronomy and navigation tools, in particular sun altitude and azimuth, are the place to compute the Sun’s position from a civil timestamp. Paste the elevation back here if the surface geometry is the remaining unknown.

Integrating cos θ over a day

Energy over a day is an integral of irradiance, not a single noon snapshot:

E = ∫ DNI(t) · cos θ(t) · dt

from sunrise to sunset, with the integrand zero when cos θ is negative. The geometric part, ∫ cos θ dt, is sometimes reported as equivalent full-sun hours on that plane. Simulation software does the sum hour by hour. This page does one instant. A noon figure of 86% does not mean the day averages 86%.

Winter days are short and the path is low: both DNI(t) and cos θ(t) are poorer for a given tilt. Summer is the reverse. Annual yield is the integral of those integrals. Orientation arguments that quote only noon on 21 June are selling a postcard, not a year.

AM1.5: 42° elevation, 48.2° zenith — not 48° elevation

Photovoltaic datasheets quote AM1.5*, the IEC 60904-3 reference spectrum. Air mass 1.5, in the simple model m = 1 / cos θz, means a zenith angle θz = arccos(1/1.5) ≈ 48.19°. Solar elevation is the complement: about 41.8°, rounded to 42°. It is not a Sun at 48° elevation.

Confusing the two swaps 6°. Elevation 48° is θz = 42°, air mass ≈ 1.35, and a different cosine on every surface. Nameplate watts (1,000 W/m², cell at 25 °C) are measured against the AM1.5 spectrum, not against “the Sun at 48°.” The 1.5 is an optical path, not an altitude angle.

IEC also specifies a 37° tilted plane and a ground albedo for the global (AM1.5G) spectrum. That is a laboratory booth, not a roof in Manchester. The incidence page here still needs the actual β, γ, latitude, date and solar time; AM1.5 is a spectral and testing convention, not a substitute for θ.

Morning versus afternoon on an east or west roof

An east-facing plane has small θ in the morning and large θ after noon; a west-facing plane does the opposite. On a clear day the two daily energy totals are close, except for a slight morning/afternoon asymmetry in air mass and typical haze. In a climate with regular afternoon storms, the east face can outperform the west on annual beam, because the hours with small θ on the west face are the hours that cloud over.

The cosine law does not know the weather. It does say that a 90° azimuth error (a due-east wall versus due south) is no longer a 5–10% story. It is a different daily shape, and a large winter penalty at mid-latitudes. Compute a June noon and a December noon on this page with the same tilt and two orientations: the December gap is the one that surprises.

A horizontal surface and a vertical wall as bookends

β = 0° is a floor, a pond, a flat roof. Then cos θ = sin α, which is also cos θz: incidence from the zenith angle, nothing else. Orientation γ drops out. The capture fraction is simply the sine of the solar elevation. At AM1.5’s 42° elevation that is sin 42° ≈ 0.67, so a horizontal patch takes about 67% of DNI before atmosphere is even discussed — and atmosphere is already inside DNI.

β = 90° is a wall. Then cos θ = cos α · cos(Az − γ). A south wall at solar noon has Az = 180° (from north) and γ = 0° in this page’s south-based surface convention after conversion, and the formula reduces to cos θ = cos α, i.e. sin of the zenith angle. Low winter Sun, high capture on the wall; high summer Sun, low capture. That is the Trombe-wall story in one line.

Glossary

Angle of incidence θ
The angle between the incoming solar ray and the normal to the surface. Captured direct power scales as cos θ. θ = 0° is square-on; θ ≥ 90° means the front face sees no beam.
Lambert’s cosine law
For a parallel beam, irradiance on a surface equals the irradiance on a facing plane times cos θ. It is the projected-area factor, equivalent to a dot product between the ray direction and the surface normal.
Trombe wall
A dark, massive wall set behind glazing on a sun-facing elevation. It stores solar heat and releases it with a delay. Winter’s low Sun favours a vertical south face; summer’s high Sun reduces the beam on that same face.
Equation of time
The difference, through the year, between apparent solar time and mean solar time, caused by orbital eccentricity and axial tilt. It reaches about ±16 minutes and, with longitude, sits between clock time and true solar time.
AM1.5
The reference solar spectrum for testing photovoltaic modules (IEC 60904-3). Air mass 1.5 corresponds to a zenith angle θz ≈ 48.2°, hence a solar elevation of about 42°, not 48°. Nameplate power is measured against that spectrum at 1,000 W/m² and 25 °C cell temperature.

Related calculators