Forty-Seven Degrees Waiting for an Almanac I Don't Have

The sextant arrived Wednesday—a Davis Mark 25, full-view plastic frame with a 4x40 monocular and a whole-horizon mirror. Not brass and mahogany. Not a museum piece. The kind of instrument you can drop on a cockpit floor and pick back up without crying, which the seller’s listing promised was tested personally during a crossing to Hawaii in 1987.

The theodolite sits on the same shelf, still in its case from last week’s triangulation session. Both instruments measure angles. Both read through vernier scales. Both require swinging the measurement to eliminate tilt error. One points at hilltops and benchmarks. The other points at the Sun.

I thought buying the sextant was the hobby. Turned out buying the sextant was step one of five.

The Almanac Problem

Took the instrument outside at 11:20 AM, sighted the Sun through the index shades, brought it down to the horizon line, read the arc: 49° 00.0’. Wrote that down. Checked the time: 11:20:17 Mountain Daylight Time. Wrote that down too.

Opened the sight reduction app on my phone, punched in the numbers, got back: “Almanac data required for calculation.”

Right. The sextant tells you the angle. The angle is useless without knowing where the Sun is supposed to be at that exact second. That information lives in the nautical almanac—a yearly publication that tabulates the celestial coordinates of the Sun, Moon, planets, and 57 navigational stars for every second of the year. The Greenwich Hour Angle, the declination, the equations of time. Pre-calculated spherical astronomy you look up instead of deriving.

I don’t have one. The 2026 edition costs $40 and ships in four days.

The Chronometer Problem

The phone says 11:20:17. But does it? MDT is six hours behind UTC, so the actual Greenwich time is 17:20:17. Probably. Unless the cell network is experiencing clock drift. Or I’ve misread the timezone offset.

Four seconds of time error equals one nautical mile of position error. That ratio is burned into maritime navigation because longitude depends on knowing when you took the sight. The Earth rotates 15° of longitude per hour. Per minute that’s 15 arcminutes—fifteen nautical miles. Per second: 0.25 nautical miles. Four seconds gets you to one.

This is why marine chronometers used to cost more than the ships carrying them. John Harrison spent forty years building H4, the first chronometer accurate enough to solve the longitude problem, and the Board of Longitude refused to pay him the full prize because they couldn’t believe a watch could do what sextants and lunar tables couldn’t.

Modern GPS time is accurate to within 40 billionths of a second. My phone syncs to it automatically. The chronometer problem is solved for anyone with a charged battery, which makes celestial navigation vastly more accurate now than it was during the entire age of sail. The irony is that GPS also makes celestial navigation obsolete, except as backup.

Commercial airline pilots still learn it. Oceanic crossings where GPS coverage is uncertain, or where electronic failure could leave you mid-Atlantic with no position fix. The stratospheric balloon work taught me that APRS beacons fail, batteries die, antennas break. Backup systems matter.

The Table Problem

Even with the almanac and a chronometer, you still need sight reduction tables. The math is spherical trigonometry—calculating your position from the angle you measured, the Sun’s tabulated position, and your estimated location. You solve for the difference between observed and calculated altitude, then plot that as a line of position on a chart.

The formula involves haversines. You can do it longhand with a calculator, but it’s slow and error-prone. Sight reduction tables pre-compute the trig for thousands of combinations of latitude, declination, and local hour angle. You look up three values, add and subtract, get an intercept distance and azimuth. Five minutes per sight instead of thirty.

Pub. 229 is six volumes. H.O. 249 is three. I have neither. The phone app does the calculation instantly, but now I’m navigating with a device that needs charging, which defeats the premise of backup navigation.

The Chart Problem

Even if I had the almanac, the chronometer discipline, and the tables, I’d need a chart to plot the line of position. A sight gives you a circle on the Earth’s surface—every point on that circle sees the Sun at the same altitude. You’re somewhere on that circle. Take another sight of a different body, get another circle. Where they intersect: your position.

The circles are enormous. A Sun sight at 47° altitude puts you on a circle roughly 4,800 kilometres in radius, centered at the Sun’s geographic position. The line you plot on the chart is actually a short tangent segment of that circle, treated as straight because you’re working at scales where the curvature is negligible.

Three sights give you a triangle. The size of the triangle tells you how accurate your measurements were. A perfect triangle collapses to a point. A sloppy one spans miles.

I don’t have a chart. The benchmark hunting work used topographic maps and survey datasheets, but those are for terrestrial coordinates. Nautical charts show depths, coastlines, hazards. Plotting sheets are blank grids you fill in yourself.

What I Actually Have

One sextant. One angle measurement I can’t reduce. A phone app that promises to do the math if I bring the other four pieces.

The theodolite measured angles to triangulate land positions, and I could close the loop in an afternoon because the reference points were fixed, surveyed, published. Celestial navigation measures angles to bodies that move continuously, requiring tables that account for orbital mechanics and Earth rotation, synchronized to time standards that didn’t exist until radio.

The precision cascade continues, but the error sources multiply: instrument accuracy (±0.1 arcminutes theoretically, ±1.5 nautical miles practically), atmospheric refraction bending starlight as it passes through air layers, temperature warping the metal frame, wave motion on a ship’s deck, the horizon itself varying with observer height and visibility.

Took another sight at 14:37. Sun altitude: 55° 00.0’. Wrote it down. Still can’t reduce it. The measurement sits in a notebook, geometrically precise and navigationally useless, waiting for an almanac that’s four days from arriving.