EN

Sight Reduction Calculator (Intercept Method)

The Sun’s lower limb has just been brought down onto the horizon in the index glass. The chronometer is stopped to the second, the micrometer read to a tenth of a minute. Those two figures — time and sextant altitude — do not hand over a position. They hand over a constraint: a line on which the vessel lies.

An assumed position is taken, the altitude the Sun ought to have there is computed, and that altitude is compared with the one actually measured. The difference, the intercept, reads in nautical miles: one minute of arc of altitude is one mile. Laid off along the body’s azimuth, it gives the point through which the line of position runs.

This calculator handles the Sun and covers both halves of the job: the sextant corrections (index error, dip, refraction, parallax, semi-diameter) and then the reduction itself. All times are Universal Time (UT/UTC), exactly as in the Nautical Almanac — not ship’s time.

NAPDPZn = 128.0°a = 6.4 NM towardsline of position ⟂ to the azimuth
Reference diagram (sample values). From the assumed position AP, lay the intercept a off along the body’s azimuth Zn — towards it if the intercept is “towards”, the opposite way if it is “away”. The line of position runs through the determined point DP, at right angles to the azimuth.

Sight, assumed position and conditions

The sight

Universal Time, not ship’s time.

Raw sextant reading: “42°18.5′” or “42.308”.

Signed: positive if it is added to the reading, negative if subtracted.

Gives the dip of the horizon: 1.76′ · √h.

Assumed position and atmosphere

“49°30.0′ N”, “49.5” or “33 52 S”.

East positive; west either negative or followed by “W”.

Refines the refraction. If no reading is available, leave 1013.

Only matters for low altitudes.

Words marked with an asterisk are unpacked in the glossary at the foot of the page — the language of the sight book, not of a tide table.

The circle of equal altitude and the geographical position

A true altitude is an angle above the horizon. At the Earth’s centre the complement z = 90° − H is the zenith distance. Every spot on the globe from which the body stands at that same altitude lies on a circle: centre at the geographical position* (the subsolar point, for the Sun), radius z. Altitude 45° means a circle 2,700 miles across. Huge. Useless as a drawn ring on a 1:200,000 chart.

That circle of equal altitude* is the geometric locus of the sight. The vessel is somewhere on it, nowhere else — provided the altitude is honest. What gets plotted is a short arc of that circle, near the dead-reckoning position. A short arc, at DR scale, looks like a straight line.

The radius is zenith distance in nautical miles: one minute of arc on the globe is one mile. A high Sun (70°) draws a tighter circle than a low one (20°). Tighter is not the same as easier to shoot.

When a short arc may be drawn as a straight LOP

The line of position* is the tangent to that circle, beside the assumed position. Over a few miles, an arc of 3,000-mile radius is indistinguishable from its chord. Hence the working plot: a straight stroke, square to the body’s azimuth, and nothing more.

The straight-line stand-in holds while the true position is within about 30 or 40 miles of the assumed one. Beyond that the curve shows, and the “line” lies by a few miles. The assumed position is slid closer and the triangle is run again. That is a sphere, not fussiness.

An intercept* of 80 miles is not a scandal. It is a signal. The DR has wandered, or the time has jumped, or the altitude was misread. At that point the LOP is not forced onto the chart. The triangle is recomputed from the foot of the perpendicular as the new assumed position. Two iterations almost always suffice.

Sight reduction by intercept: Ho, Hc, towards or away

Marcq St Hilaire* does not hunt for the centre of the circle. Two altitudes are compared at the same instant, from the same supposed position. One, Ho*, comes from the sextant after corrections. The other, Hc*, comes from the PZX triangle, computed from the assumed position, the declination and the local hour angle.

22 June, 08:18 UT, south of the Needles. Assumed position 50° 32′ N, 001° 35′ W. Sun sight. Hs = 41° 07.4′. After the correction chain, Ho = 41° 19.1′. For that AP and that second, the triangle yields Hc = 41° 15.8′ and Zn = 098°. Intercept +3.3′. 3.3 miles are laid off towards the Sun, then the perpendicular to Zn. That is the LOP.

sin Hc = sin φa · sin δ + cos φa · cos δ · cos LHA

a = (Ho − Hc) in minutes of arc = nautical miles, laid off along azimuth Zn

The sign is a displacement. a > 0, towards: the observed altitude exceeds the computed one, so the ship is closer to the GP than the DR claimed. a < 0, away: farther off, and the intercept is laid off opposite the body. The form does Hc and a. The pencil on the chart does the rest. One LOP is not a fix. It is a constraint.

Zn* is not a compass bearing. It is the astronomical azimuth, from true north eastward, 0° to 360°. The LOP itself bears Zn + 90° / Zn − 90°. It is not drawn “towards the Sun”. It is drawn across the Sun’s direction.

Altitude corrections in the order they belong

The raw sextant reading, Hs*, is not a true altitude. Five terms sit between the micrometer and the sky. Each meets the altitude as it stands at that step. Skip one, reverse two, clap parallax* on Vega: the fix walks off three miles.

  1. Index error* — the instrument. Taken in the morning, signed, written down.
  2. Dip* — the eye is not at sea level. The apparent horizon drops. 1.76′ · √h, h in metres. Eye 2.5 m: 2.8′. Eye 9 m, a ship’s bridge: 5.3′.
  3. Refraction* — air lifts the body. About 34′ at the horizon, 1′ near 45°, almost nothing in the zenith. Bennett’s formula, corrected for pressure and temperature, is enough at sea. Below 15° of altitude it starts to care about the day’s weather.
  4. Parallax — the observation is not from the Earth’s centre. Moon, up to 1°. Sun, a fraction of a minute (around 0.15′). A star, zero.
  5. Semi-diameter* — the limb was shot, not the centre. Sun and Moon only. Lower limb: add. Upper limb: subtract. Centre of the body (rare, a veiled Sun): nothing.

Ho = Hs + index corr. − dip − refraction + parallax ± semi-diameter

The Sun’s semi-diameter is about 16′. Leave it out and the LOP is sixteen miles adrift. Shoot the upper limb “because the lower is in haze” and still apply the lower-limb sign: thirty-two miles. The sight book records the limb, not only the altitude. Apparent altitude Ha, after index and dip, is the figure Bennett acts on — not Hs.

Index error, on and off the arc

The horizon is brought into coincidence, both mirrors in line. The offset, in minutes, is index. It moves: temperature, a knock in the locker, the tangent screw pinched too hard the night before. Two minutes of index left on the shelf is two miles. It is taken in the morning, taken again at noon, written beside the round — not kept “in the head”.

On the arc or off the arc — the maker’s sign convention — decides the sign. Get the sign wrong once, and every LOP of the day sets off the same way. The horizon test remains the simplest. A sextant that “eats” 4′ from one day to the next is not tired. It is out of adjustment, or badly stowed. Before the DR is blamed, the index is taken again.

Plotting on a Mercator chart

From the assumed position a line is drawn in the direction of Zn (towards) or Zn + 180° (away). The intercept is stepped off in nautical miles. Through that point the LOP is drawn square to the azimuth. On a Mercator chart a mile of latitude is a minute of latitude; a mile of longitude is a minute of longitude divided by the cosine of latitude. Near 50° N that factor is about 0.64. Mixing the two scales is a classic way to put a good sight in the wrong place.

Rather than scale-fiddle, many navigators plot from a convenient whole-degree AP even when the DR is not on it — the tables’ habit, carried onto a computed page. The intercept then absorbs the difference. The LOP is the same line. The arithmetic is cleaner.

Crossing LOPs and reading the cocked hat

One line: the ship is somewhere on it. Two lines: their cross is the observed position — if the azimuths are not too close. 30° apart, a decent cocked hat later. 10°, a long corridor, pretty on the chart, useless for a landfall. Three lines, a triangle of error, the real luxury.

Two Suns an hour apart, azimuths too matey, will not close. A Sun and a planet, or the Sun and Polaris in the evening, will. The dusk ideal is three stars whose azimuths sit around the horizon. The triangle that results is not a failure. It is the measure of the uncertainty.

The convenient corner of the triangle is not the one chosen for the course. The centre is taken, or the sights are weighted (clearer horizon to the west, higher star to the south). A three-mile triangle in fair weather is an honest fix. A fifteen-mile triangle says that one of the three sights, or the time, or the identification, has cheated.

Advancing an LOP for a running fix

Singlehanded, two bodies do not come at once. One is shot at 08:18, DR is run until 14:05, the other is shot. The first LOP is advanced: slid along the ground track by the distance run. The cross happens at 14:05, with an 08:18 line that has travelled.

DR is the weak link. Stream misread, log fouled, a log that “eats” 8 % in a head sea, and the advanced line arrives crooked. That is known. It is done all the same, failing a second observer, and without a second body in a daylight sky.

Advancing is not a vague slide. It is a parallel shift, course and distance of the DR between the two times. Ground track 248°, 32 miles run: every point of the old LOP moves 32 miles on 248°. The new line, parallel, cuts the second. If the DR has lied by 4 miles, the observed position lies by 4 miles — in the direction of the DR, not of the body. For all that, a 4-mile lie is often more use than DR alone. The second sight at least pins a direction.

The noon sight: latitude from meridian altitude

At apparent noon the body is on the meridian, azimuth 180° or 000°. Altitude gives latitude, almost no azimuth, almost no tight time. φ = 90° − Ho ± δ, according as the Sun culminates south or north of the zenith. It is the simplest celestial fix, and the one most yachts still take.

φ = 90° − Ho + δ   (Sun to the south, northern hemisphere, the usual case)

“Almost no time” has a limit. Miss the culmination by five minutes near the solstice and the altitude barely notices: the curve is flat. Near equinox, a little more. A string is taken around the peak, the highest kept. There is no need to time the second of apparent noon if five readings straddle the top.

A noon sight does not give longitude. It gives a parallel. To close the fix, another LOP is wanted — a morning sight advanced, a dusk planet, or a longitude by DR that nobody pretends is good to a mile. Apparent noon is not noon on the clock. The equation of time, plus longitude, shifts the passage by a quarter of an hour, sometimes more. Shooting at 12:00 zone time “because it is noon” is shooting at the wrong moment.

A round of sights rather than a single shot

Five shots in three minutes, time in the middle, the wild one thrown out, the rest averaged. Swell bounces the limb. One Hs is a snapshot. A round is a measurement.

Feet are not shifted between shots. The horizon is the same. The eye is the same. The shade is the same. A recorder calling the time while the observer holds the contact — the classic pair on a bridge — beats a single-hander who drops the sextant to look at the watch. Failing a second pair of hands, a chronometer is started on contact and read afterwards.

Honest scatter, in summer weather, runs 0.3′ to 0.8′. Beyond that, something is wrong: a moving horizon, a stiff arm, a Sun that breaks through and veils again. A round spread over 3′ is not “saved” by keeping the middle. It is retaken, or another body is used.

Chronometer time: one minute of clock, fifteen miles of longitude

The sky turns 15° an hour, 15′ a minute of time. One minute of clock too late is fifteen minutes of local hour angle* — up to fifteen miles of longitude at the equator, less in high latitudes, always too much. The chronometer, or GPS time, is checked. A sight without a second is not a sight.

Ship’s time is not almanac time. Zone, summer time, “the clocks went forward at six”: so many traps. The tables, and this page, speak UT*. The sight watch is compared with UT before the round, and after. The error, the chronometer rate, is logged like the index.

Altitude is minded to a tenth. Time, to a second. Both spend the same miles on the chart. Even so, the contact is no longer “read the sextant, then the watch”. It is the other way round, or both at once. The book carries the time first, the altitude second. A micrometer can be read again. A lost second cannot.

HO 249, HO 229 and NP 401

The American HO 249* (Sight Reduction Tables for Air Navigation) and HO 229*, and the British cousin NP 401, gave Hc and Zn for a whole degree of AP, from LHA, latitude and declination. Interpolation did the rest. This page solves the triangle straight. It does not replace the almanac (δ, GHA), or the sextant, or the chart. It replaces the page-turning and the interpolation.

The AP of the tables is not any DR. Latitude is taken whole, and longitude is chosen so that LHA falls on a whole degree. That simplifies the entry. It does not improve the DR. GPS off, the log’s DR remains what it is: a starting point, not a result.

At sea a wet book of tables is still a book. A phone overboard is no longer a calculator. A passage is not sailed with only one method. Tables, pencil, this page, the triangle by hand if need be: three roads to the same Hc. The one that survived the spray wins.

Swell, an uncertain horizon, abnormal refraction

A slamming horizon, the limb is shot high or low with the trough. A misty horizon, a line is invented. Very cold air over warm water, refraction is not the table. All the LOPs sit the same side of truth, and the ship “knows” where it is, three miles out.

Abnormal refraction is not cured by a coefficient. It is suspected: mirage, a fuzzy horizon, a sea–air temperature gap of more than a few degrees, low sights that refuse to agree with high ones. Altitudes under 15° are set aside. High bodies are kept. The day’s fix is allowed to be uglier than yesterday’s.

Swell is fought with the round, not with a magic “shoot in the trough”. The eye looks for the true horizon, not the top of the swell. Three sights high on the same side often mean the horizon was taken too high, not that the DR has slipped.

An LOP is not a GPS fix. It is a constraint, good to a mile or three in fair weather, worse in a seaway. A rock is not shaved with it.

The assumed position is not a GPS fix

Marcq wants a supposed position, not the true one. GPS can be dropped into the AP; the intercept should then sit near zero if the sight is honest — a sextant check, not navigation. Real nav, the AP comes from DR. GPS off, that is what remains.

Dropping GPS into the form has one honest use: checking the instrument. Intercept 0.3 mile, azimuth coherent, index checked: the sextant is telling the truth. Intercept 12 miles with the yacht still on the pontoon: index, time or limb has cheated. That is a bench test, not a method of passage-making.

What the sight book has to carry

A reduction exists only if it can be read back. For each sight the book carries, at least: date, UT to the second, Hs, limb, index of the moment, height of eye, pressure and temperature if known, assumed position, then Ho, Hc, a, Zn. Without that, an LOP that “will not sit” cannot be unpicked.

Horizon state (sharp, hazy, choppy), the shade used, the observer’s name, go in as well. Those notes look like folklore. In the evening they separate the 07:50 sight (clean horizon) from the 16:20 one (squall, Sun in the wake). After all, the book outlives the phone. A pencil log, even wet, can still be read by lamplight.

Glossary

Geographical position (GP)
The point on the Earth where the body is in the zenith at the instant of the sight. Centre of the circle of equal altitude. For the Sun, the subsolar point.
Circle of equal altitude
The locus of all positions from which a body is seen at the same true altitude. Its radius is the zenith distance, 90° − H, in nautical miles. The LOP is the local tangent.
Line of position (LOP)
The straight-line stand-in for a short arc of the circle of equal altitude. Drawn through the intercept terminal, at right angles to the body’s azimuth.
Intercept
The difference a = Ho − Hc, in minutes of arc and therefore in nautical miles. Positive (towards): laid off towards the body. Negative (away): laid off the opposite way.
Marcq St Hilaire
The intercept method of sight reduction, after the French naval officer Adolphe Marcq de Saint-Hilaire (1875). The altitude that should be seen from the assumed position is computed, compared with the observed altitude, and the difference laid off along the azimuth.
Ho (observed altitude)
The body’s altitude above the true horizon after every sextant correction. The quantity compared with Hc. French practice writes Hv.
Hc (computed altitude)
The altitude the body should have, as seen from the assumed position, at the instant of the sight. It comes out of the PZX triangle (latitude, declination, LHA).
Zn (azimuth)
The body’s true direction, from north eastward, 0° to 360°. The LOP is perpendicular to it. Not a compass bearing.
Hs (sextant altitude)
The raw reading on the arc and micrometer at the moment of contact, before any correction.
Index error
The sextant’s zero offset, in minutes of arc, found by bringing both mirrors onto the horizon. Signed on or off the arc according to the instrument. It moves: it is retaken every round.
Dip
The depression of the apparent horizon caused by height of eye above the water. Usual sea formula: 1.76′ · √h, h in metres. Always subtractive.
Refraction
The apparent lifting of the body by the atmosphere. Greatest near the horizon (~34′), negligible in the zenith. This page uses Bennett’s formula, corrected for pressure and temperature.
Parallax in altitude
The difference between the body’s direction as seen from the surface and as seen from the Earth’s centre. Large for the Moon, small for the Sun (~0.15′), nil for stars.
Semi-diameter
The apparent radius of the disc, Sun or Moon. A limb is shot, not the centre: the semi-diameter is added for the lower limb, subtracted for the upper. About 16′ for the Sun.
LHA (local hour angle)
The angle at the pole between the observer’s meridian and the body’s meridian, counted west from 0° to 360°. Equivalent to the French AHl, which is often signed (negative before noon).
UT (Universal Time)
The time of the almanac and of this page, UTC. Not ship’s time, not summer time. A sight dated in zone time without conversion is a wasted sight.
HO 249
Sight Reduction Tables for Air Navigation, US Pub. No. 249, giving Hc and Zn by latitude, declination and LHA. British cousin: NP 401. Designed for air work, widely used at sea for their compactness.
HO 229
Sight Reduction Tables for Marine Navigation, denser than HO 249, to a tenth of a minute, intended for the ship. Finer interpolation, more bulk.

Related calculators