USNO 1999 Benchmark Verification

Algorithmic integrity and mathematical correspondence with the U.S. Naval Observatory multi-sight position fix standard.

Executive Summary

When relying on an electronic celestial navigation aid far offshore, a navigator needs absolute confidence that the underlying algorithms are mathematically sound, rigorous, and verified against international standards.

To demonstrate this precision, Vincenta Ltd benchmarked Vincenta’s Celestial Running Fix Engine was benchmarked against the official 1999 U.S. Naval Observatory (USNO) multi-sight test problem published jointly by the USNO and Her Majesty’s Nautical Almanac Office (HMNAO) in The Nautical Almanac 1999 (pages 282–283).

The result: Vincenta reproduces the USNO published benchmark step-for-step, converging on the exact same position fix within 0.01 to 0.02 arcminutes (∼15 to 30 meters)—a tiny variance fully accounted for by Vincenta’s high-precision 64-bit floating-point arithmetic and WGS-84 ellipsoidal geometry.

The Benchmark Problem Setup

The USNO test problem represents a real-world offshore scenario: a vessel steaming at 20.0 knots on a course of 325° True, taking a round of three star sights across a 31-minute observation window on 4 July 1999.

  • Initial DR Estimate at Fix Time (21:00:00 UT): N 32° 00.00', W 15° 00.00'
  • Sight 1 — Regulus (No. 26): 20h 39m 23s UT — Observed Altitude (Ho): 27° 14.0' (27.2333°)
  • Sight 2 — Antares (No. 42): 20h 45m 47s UT — Observed Altitude (Ho): 25° 57.2' (25.9536°)
  • Sight 3 — Kochab (No. 40): 21h 10m 34s UT — Observed Altitude (Ho): 47° 28.0' (47.4664°)

Because the observations span 31 minutes while the vessel covers over 10 nautical miles, the observations must be motion-compensated to the 21h 00m 00s UT fix time before intersecting the Lines of Position (LOPs).

Side-by-Side Numerical Comparison

The table below compares the intermediate iterations and final fix coordinates published in The Nautical Almanac 1999 against the output generated by Vincenta on the TI-Nspire™ CX handheld.

Computation Phase Official USNO 1999 Publication Vincenta Handheld Output Variance / Agreement
Initial DR Baseline N32° 00.00', W15° 00.00' N32° 00.00', W15° 00.00' Exact Match
Iteration 1 Position N31° 37.16', W15° 01.13' N31° 37.14', W15° 01.15' 0.02' Lat / 0.02' Lon (∼30 m)
Iteration 1 Shift (d) 22.9 NM 22.88 NM 0.02 NM (∼36 m)
Iteration 2 (Final Fix) N31° 37.15', W15° 01.15' N31° 37.13', W15° 01.16' 0.02' Lat / 0.01' Lon (∼15 m)
Iteration 2 Shift (d) 0.02 NM 0.02 NM Exact Match (0.00 NM)

How the Algorithm Works Under the Hood

Vincenta’s running fix engine implements the official USNO multi-body algorithm across three integrated stages:

  1. Motion Compensation (Time Transposition): Advances or retards each observation along the vessel’s course over ground (325°) and speed (20.0 kt) to the common fix time (21h 00m 00s UT) using WGS-84 rhumb-line vector geometry.
  2. Simultaneous Least-Squares Engine: Evaluates all available sights simultaneously using the USNO matrix system of normal equations (A, B, C, D, E, G), incorporating dynamic sight-weighting factors.
  3. Iterative Convergence Control: Per the USNO standard, if the initial position shift exceeds 20 nautical miles, the algorithm re-reduces the sights from the improved position and runs a second iteration until shift distance ≤ 20 NM.

Understanding the Micro-Variance (∼15 Meters)

Vincenta’s final latitude (N31° 37.13') differs from the printed USNO page (N31° 37.15') by just 0.02 arcminutes (∼36 meters). This minute difference favors Vincenta and is explained by two factors:

  • Unrounded Arithmetic: The published USNO paper truncates intermediate terms to 4 decimal places for manual illustration. Vincenta carries unrounded double-precision numbers throughout the entire calculation chain.
  • WGS-84 Ellipsoidal Model: USNO's paper formulas use a spherical Earth model with a simple cos(Lat) scaling factor. Vincenta calculates rhumb line transpositions using the full WGS-84 reference ellipsoid.

Printable Technical Report

A printable PDF copy of the complete USNO 1999 Benchmark Technical Verification Report is available below.

Download USNO 1999 Benchmark Report (PDF)

Navigational Confidence at Sea

Vincenta’s performance on the USNO 1999 benchmark confirms that its running fix engine is a fully verified, highly accurate digital execution of the premier USNO multi-sight algorithm.



Purchase Vincenta