# Atmospheric refraction

Not to be confused with Atmospheric diffraction.
Diagram showing displacement of the Sun's image at sunrise and sunset

Atmospheric refraction is the deviation of light or other electromagnetic wave from a straight line as it passes through the atmosphere due to the variation in air density as a function of altitude. This refraction is due to the velocity of light through air decreasing (the index of refraction increases) with increased density. Atmospheric refraction near the ground produces mirages and can make distant objects appear to shimmer or ripple, elevated or lowered, stretched or shortened with no mirage involved. The term also applies to the refraction of sound. Atmospheric refraction is considered in measuring the position of both astronomical and terrestrial objects

Astronomical or celestial refraction causes astronomical objects to appear higher in the sky than they are in reality. Terrestrial refraction usually causes terrestrial objects to appear higher than they really are, although in the afternoon when the air near the ground is heated, the rays can curve upward making objects appear lower than they really are.

Refraction not only affects lightrays but all electromagnetic radiation, although in varying degrees (see dispersion in optics). For example in visible light, blue is more affected than red. This may cause astronomical objects to be spread out into a spectrum in high-resolution images.

Whenever possible, astronomers will schedule their observations around the time of culmination of an object when it is highest in the sky. Likewise sailors will never shoot a star which is not at least 20° or more above the horizon. If observations close to the horizon cannot be avoided, it is possible to equip a telescope with control systems to compensate for the shift caused by the refraction. If the dispersion is a problem too, (in case of broadband high-resolution observations) atmospheric refraction correctors can be employed as well (made from pairs of rotating glass prisms). But as the amount of atmospheric refraction is a function of the temperature gradient, the temperature, pressure, and humidity (the amount of water vapour is especially important at mid-infrared wavelengths) the amount of effort needed for a successful compensation can be prohibitive. Surveyors, on the other hand, will often schedule their observations in the afternoon when the magnitude of refraction is minimum.

Atmospheric refraction becomes more severe when there are strong temperature gradients, and refraction is not uniform when the atmosphere is inhomogeneous, as when there is turbulence in the air. This is the cause of twinkling of the stars and various deformations of the shape of the sun at sunset and sunrise.

## Astronomical refraction

Astronomical refraction deals with the angular position of celestial bodies, their appearance as a point source, and through differential refraction, the shape of extended bodies such as the Sun and Moon.[1]

### Values

Atmospheric refraction of the light from a star is zero in the zenith, less than 1′ (one arc-minute) at 45° apparent altitude, and still only 5.3′ at 10° altitude; it quickly increases as altitude decreases, reaching 9.9′ at 5° altitude, 18.4′ at 2° altitude, and 35.4′ at the horizon;[2] all values are for 10 °C and 101.3 kPa in the visible part of the spectrum.

On the horizon refraction is slightly greater than the apparent diameter of the Sun, so when the bottom of the sun's disc appears to touch the horizon, the sun's true altitude is negative. If the atmosphere suddenly vanished, the sun would too. By convention, sunrise and sunset refer to times at which the Sun’s upper limb appears on or disappears from the horizon and the standard value for the Sun’s true altitude is -51.4`: −35.4′ for the refraction and −16′ for the Sun’s semi-diameter. The altitude of a celestial body is normally given for the center of the body’s disc. In the case of the Moon, additional corrections are needed for the Moon’s horizontal parallax and its apparent semi-diameter; both vary with the Earth–Moon distance.

Day-to-day variations in the weather will affect the exact times of sunrise and sunset[3] as well as moon-rise and moon-set, and for that reason it generally is not meaningful to give rise and set times to greater precision than the nearest minute.[4] More precise calculations can be useful for determining day-to-day changes in rise and set times that would occur with the standard value for refraction[note 1] if it is understood that actual changes may differ because of unpredictable variations in refraction.

Because atmospheric refraction is 34′ on the horizon, but only 29′ at 0.5° above it, the setting or rising sun seems to be flattened by about 5′ (about 1/6 of its apparent diameter).

The light from distant objects on the earth is refracted too; the straight line from your eye to a distant mountain might be blocked by a closer hill, but the actual light path may curve enough to make the distant peak visible. A reasonable first guess: a mountain's apparent altitude at your eye (in degrees) will exceed its true altitude by its distance in kilometers divided by 1500. This assumes a fairly horizontal line of sight and ordinary air density; if the mountain is very high (so much of the sightline is in thinner air) divide by 1600 instead.

### Calculating refraction

Plot of refraction vs. altitude using Bennett’s 1982 formula

Young[6] distinguished three regions where different methods for calculating refraction were applicable. In the upper portion of the sky, with a zenith distance of less than 60° (or an altitude over 30°), various simple refraction formulas based on the index of refraction (and hence on the temperature, pressure, and humidity) at the observer are adequate. Between 30° and 5° of the horizon, the temperature gradient becomes the dominant factor, and formulas employing the average temperature gradient are needed. Closer to the horizon, details of how the local temperature gradient changes with height become important, and rigorous calculations of refraction require numerical integration, using a method such as that of Auer and Standish.[7]

Bennett[8] developed a simple empirical formula for calculating refraction from the apparent altitude, using the algorithm of Garfinkel[9] as the reference; if ha is the apparent altitude in degrees, refraction R in arcminutes is given by

$R = \cot \left ( h_\mathrm{a} + \frac {7.31} {h_\mathrm{a} + 4.4} \right ) \,;$

the formula is accurate to within 0.07′ for the altitude range 0°–90°.[4] Sæmundsson[10] developed a formula for determining refraction from true altitude; if h is the true altitude in degrees, refraction R in arcminutes is given by

$R = 1.02 \cot \left ( h + \frac {10.3} {h + 5.11} \right ) \,;$

the formula is consistent with Bennett’s to within 0.1′. Both formulas assume an atmospheric pressure of 101.0 kPa and a temperature of 10 °C; for different pressure P and temperature T, refraction calculated from these formulas is multiplied by[4]

$\frac {P} {101} \, \frac {283} {273 + T}$

Refraction increases approximately 1% for every 0.9 kPa increase in pressure, and decreases approximately 1% for every 0.9 kPa decrease in pressure. Similarly, refraction increases approximately 1% for every 3 °C decrease in temperature, and decreases approximately 1% for every 3 °C increase in temperature.

### Random refraction effects

An animated image of the Moon's surface showing the effects of Earth's atmosphere on the view

Turbulence in the atmosphere magnifies and de-magnifies star images, making them appear brighter and fainter on a time-scale of milliseconds. The slowest components of these fluctuations are visible as twinkling (also called "scintillation").

Turbulence also causes small random motions of the star image, and produces rapid changes in its structure. These effects are not visible to the naked eye, but are easily seen even in small telescopes. They are called "seeing" by astronomers.

## Terrestrial refraction

Terrestrial refraction deals with the apparent angular position and measured distance of terrestrial bodies. It is of special concern for the production of precise maps and surveys.[11]

## Notes

1. ^ For an example see Meeus 2002[5]

## References

1. ^ Bomford, Guy (1980), Geodesy (4 ed.), Oxford: Oxford University Press, pp. 282–284, ISBN 0-19-851946-X
2. ^ Allen, C.W. (1976). Astrophysical quantities (3rd ed. 1973, Repr. with corrections 1976. ed.). London: Athelone Press. p. 125. ISBN 0-485-11150-0.
3. ^ Schaefer, Bradley E.; Liller, William (1990). "Refraction near the horizon". Publications of the Astronomical Society of the Pacific 102: 796–805. Bibcode:1990PASP..102..796S. doi:10.1086/132705.
4. ^ a b c Meeus, Jean (1991). Astronomical algorithms (1st English ed. ed.). Richmond, Va.: Willmann-Bell. pp. 102–103. ISBN 0-943396-35-2.
5. ^ Meeus, Jean (2002). [Mathematical astronomy morsels] (1st English ed. ed.). Richmond, Va.: Willmann-Bell. p. 315. ISBN 0-943396-74-3.
6. ^ Young, Andrew T. (2006). "Understanding Astronomical Refraction". The Observatory 126: 82–115. Bibcode:2006obs...126...82y.
7. ^ Auer, Lawrence H.; Standish, E. Myles (2000). "Astronomical Refraction: Computation for All Zenith Angles". Astronomical Journal 119 (5): 2472–2474. Bibcode:2000AJ....119.2472A. doi:10.1086/301325.
8. ^ Bennett, G.G. (1982). "The Calculation of Astronomical Refraction in Marine Navigation". Journal of Navigation 35: 255–259. Bibcode:1982JNav...35..255B. doi:10.1017/S0373463300022037.
9. ^ Garfinkel, B. (1967). "Astronomical Refraction in a Polytropic Atmosphere". Astronomical Journal 72: 235–254. Bibcode:1967AJ.....72..235G. doi:10.1086/110225.
10. ^ Sæmundsson, Þorsteinn (1986). "Astronomical Refraction". Sky and Telescope 72: 70. Bibcode:1986S&T....72...70S.
11. ^ Bomford, Guy (1980), Geodesy (4 ed.), Oxford: Oxford University Press, pp. 42–48, 233–243, ISBN 0-19-851946-X