Sextant altitude corrections, step by step
Turns a sextant altitude into an observed altitude and shows every step: index correction, dip of the horizon, refraction for the air you measured, semidiameter for the limb you touched, and parallax
Planning and education aid. Not for primary navigation. Full disclaimer
the sight is worth 51°35.5', 7.1 minutes more than the sextant read.
| Step | Value |
|---|---|
| Index correction | -2.0 arcmin |
| Dip | -6.0 arcmin |
| Refraction | -0.8 arcmin |
| Semidiameter | 15.8 arcmin |
| Parallax | 0.1 arcmin |
- Observed altitude
- Apparent altitude
- Total correction
- Dip of the horizon
- Refraction
- Semidiameter
Provenance
- Computed by
- navigation.celestial.altitude-correction 1.0.0, core 0.1.0
- Model
- Apparent altitude ha = hs + index correction - dip, with dip 1.76' times the square root of the height of eye in metres; refraction by Bennett (1982), 1/tan(ha + 7.31/(ha + 4.4)) arcminutes, scaled by the pressure over 1010 hPa and 283 K over the air temperature; semidiameter from the body's computed distance for the Sun and from asin(0.2725076 sin HP) for the Moon, added for a lower limb and taken off for an upper; parallax in altitude HP cos ha
- Accuracy
- Within 0.15′ of the six worked examples in Bowditch chapter 16 above 5° of apparent altitude, and within 0.25′ of the two below it; refraction within 0.02′ of the US Naval Observatory from 1.4° to 89° of apparent altitude
- Reference data
- version IERS Bulletin C 72 (2026-07-06) of a reference dataset
- Notes
- None
- Cites
- National Geospatial-Intelligence Agency, The American Practical Navigator (Bowditch), NGA Pub. No. 9, Volume I; Bennett, G. G., The Journal of Navigation (Royal Institute of Navigation), The Calculation of Astronomical Refraction in Marine Navigation; Bureau des Longitudes (now IMCCE), distributed by CDS Strasbourg, VSOP87 planetary theories (Bretagnon and Francou 1988), VizieR catalog VI/81; Bureau des Longitudes (now IMCCE), distributed by CDS Strasbourg, ELP 2000-82B lunar solution (Chapront-Touze and Chapront 1983, 1988), VizieR catalog VI/79; US Naval Observatory, Astronomical Applications Department, USNO Astronomical Applications API (celestial navigation and sidereal time)
Something look off?
How we got thisFormula, worked example, sources, and proof
Model: Apparent altitude ha = hs + index correction - dip, with dip 1.76' times the square root of the height of eye in metres; refraction by Bennett (1982), 1/tan(ha + 7.31/(ha + 4.4)) arcminutes, scaled by the pressure over 1010 hPa and 283 K over the air temperature; semidiameter from the body's computed distance for the Sun and from asin(0.2725076 sin HP) for the Moon, added for a lower limb and taken off for an upper; parallax in altitude HP cos ha
Show your work
Apparent altitude
ha = hs + index correction - dip, dip 1.76' root metres51°28.4' with eye 11.58 m= ha 51°20.4'Refraction
Bennett 1982, scaled for the pressure and temperature1010 hPa at 10 degrees Celsius= 0.8 minutesLimb and parallax
Semidiameter for the limb, then HP cos halower limb, Sun= +15.8 and +0.1 minutesObserved altitude
Everything added to the sextant altitude51°28.4' +7.1 minutes= the sight is worth 51°35.5', 7.1 minutes more than the sextant read
The same steps an agent gets from the MCP server with explain: true.
Accuracy: Within 0.15′ of the six worked examples in Bowditch chapter 16 above 5° of apparent altitude, and within 0.25′ of the two below it; refraction within 0.02′ of the US Naval Observatory from 1.4° to 89° of apparent altitude
When to use this: Use this between the sextant and the sight reduction: give the altitude you read, the index correction, your height of eye, which limb you brought to the horizon, and the air temperature and pressure if you measured them. It returns the observed altitude with every correction shown, so the work form can be checked line by line.
Limitations: It is written for a marine sextant on the sea horizon. Artificial horizons, back sights, and bubble sextants need their own corrections, which are not here. Refraction below about 5 degrees of altitude is variable and not reliably modelled by any formula, so treat a low sight as approximate. Agreement with the Nautical Almanac's tables is a check, not a promise; that publication governs.
Worked example: The Sun's lower limb at 51 degrees from 38 feet, 2024-06-02. Source: Bowditch 2024 §1618 Example 1: hs 51°28.4', index correction -2.0', dip -6.0', ha 51°20.4', total correction +15.2', Ho 51°35.6'. It is golden test vector v001, and every build checks the tool still gives its answer within its tolerance.
You enter
- Body
- Sun
- DUT1 (UT1 - UTC)
- 0 s
- Height of eye
- 38 ft
- Sextant altitude
- 51 28.4
- Index correction
- -2.0 arcmin
- Limb observed
- lower
- UTC time
- 2024-06-02T11:00:00Z
You get
- Observed altitude
- the sight is worth 51°35.5', 7.1 minutes more than the sextant read
- Observed altitude
- 51.5912°
- Apparent altitude
- 51.3402°
- Total correction
- 7.1′
- Dip of the horizon
- -6′
- Refraction
- -0.8′
- Semidiameter
- 15.8′
- Parallax in altitude
- 0.1′
- Index correction
- -2′
Review: Not yet independently reviewed by a geodesist.
Last verified: 2026-09-19, when a maintainer last confirmed this tool's sources at the issuer. See the sources ledger.
Status: version 1.0.0, core 0.1.0. See this tool in the verification report.
Changes
- 2026-11-06, changed: Eight marine and celestial tools are now stable: sextant altitude corrections, sight reduction, set and drift, plane sailing, tide datum conversion, tide prediction, the moon's position, and moonrise and moonset. Nothing they compute changed. Each was already checked against a published source rather than against itself: Bowditch's own worked examples for the sights, the sailings, and the current triangle, NOAA's published tide predictions and station datums for the tides, and the US Naval Observatory's computed data for the moon. What the stable mark adds is that the build now holds them to the whole bar on each commit. Their pages no longer carry the experimental warning. The navigation MCP toolset grows to 37 tools, still under its cap. Changelog
- 2026-10-28, changed: Two drawings and two explainers for the celestial and moon tools. The sextant altitude corrections now draw as the work form a navigator writes down the margin: the altitude read off the arc at the top, each correction stepping it left or right by its own size to scale, and the observed altitude at the bottom. Moonrise and moonset now draw a sky plot with the moon's path and the sun's behind it, north up and the horizon at the rim, with rise and set marked where the moon crosses it; the tool returns both paths for the day to make that possible. Two new explainers: how a sextant sight becomes a line of position, and how much light the moon will give you, which covers why a half moon is far dimmer than half as bright and why a calendar date can hold two moonrises or none. Changelog
- 2026-10-27, added: New, both experimental: sextant altitude corrections and sight reduction. The first turns the altitude you read off the arc into the observed altitude, showing the index correction, the dip of the horizon, refraction for the air you measured, the semidiameter of the limb you touched, and parallax, one line at a time. The second reduces that sight to a line of position by the intercept method: the body's hour angle and declination for the second of the sight, the computed altitude and true azimuth at your assumed position with the quadrant rule stated, and how far to shift the line toward or away from the body, with the line itself drawn on the chart. The Pub. No. 229 assumed-position rule is one click away. Every worked example in Bowditch chapter 16 and three columns of its chapter 20 four-star fix are reproduced, and both tools are checked against the US Naval Observatory's own data at 148 body-instants from 1995 to 2045. The Nautical Almanac still governs. Changelog
Checked against: 35 golden test vectors (download the test vectors, each with its source and tolerance). See how results are checked and every source.
Sources
- The American Practical Navigator (Bowditch), NGA Pub. No. 9, Volume I, National Geospatial-Intelligence Agency, 2024 edition. Chapter 16, Sextant Altitude Corrections (§1604 dip, §1605 refraction, §1608 semidiameter, §1611 parallax, §1612 to §1613 the summary and order; §1618 to §1622, the worked examples).
- The Calculation of Astronomical Refraction in Marine Navigation, Bennett, G. G., The Journal of Navigation (Royal Institute of Navigation), Vol. 35, No. 2. pp. 255-259, the formula R = cot(h + 7.31/(h + 4.4)).
- VSOP87 planetary theories (Bretagnon and Francou 1988), VizieR catalog VI/81, Bureau des Longitudes (now IMCCE), distributed by CDS Strasbourg, VSOP87D, heliocentric ecliptic of date. Files VSOP87D.ear, .ven, .mar, .jup, .sat; Astronomy and Astrophysics 202, 309.
- ELP 2000-82B lunar solution (Chapront-Touze and Chapront 1983, 1988), VizieR catalog VI/79, Bureau des Longitudes (now IMCCE), distributed by CDS Strasbourg, ELP 2000-82B, constants fitted to DE200/LE200. Files ELP1 to ELP36 and the authors' subroutine ELP82B (elp82b.f).
- USNO Astronomical Applications API (celestial navigation and sidereal time), US Naval Observatory, Astronomical Applications Department, API 4.0.1, celnav endpoint. Celestial navigation data for assumed position and time: GHA and declination of the Sun, Moon, planets, Aries, and the navigational stars.
Terms
- DUT1: UT1 minus UTC
- The difference between Earth-rotation time and UTC, always under 0.9 s, broadcast so users can recover UT1. Source: IERS Bulletin C (leap seconds)
- HP: horizontal parallax
- The angle the Earth's radius covers as seen from a body with that body on the horizon; the Moon's is about a degree, so a sight of it must be corrected for where you stand. Source: The American Practical Navigator (Bowditch), NGA Pub. No. 9
- IERS: International Earth Rotation and Reference Systems Service
- The service that maintains the ITRF and Earth orientation data, and announces leap seconds in Bulletin C. Source: IERS Bulletin C (leap seconds)
- SD: semidiameter
- Half the apparent width of the Sun or Moon, about 16 minutes of arc; a sextant touches a limb, so this correction takes the sight to the center. Source: The American Practical Navigator (Bowditch), NGA Pub. No. 9
- UT1: Universal Time 1
- Time kept by Earth's actual rotation. It drifts against atomic time, and UTC is kept within 0.9 s of it. Source: IERS Bulletin C (leap seconds)
- UTC: Coordinated Universal Time
- The world's civil time standard, kept within 0.9 s of Earth's rotation by leap seconds. Aviation calls it Zulu time, written with a Z. Source: IERS Bulletin C (leap seconds)
Learn the concept: How a sextant sight becomes a line of position