Discrepancies in TICON constituent definitions

[bkeepers]

The TICON-4 User Manual includes constituent definitions used for generating the harmonics. Most of the definitions match what is used internally in @neaps/tide-predictor, but there are 6 discrepancies:

1. SA (Sa) — Doodson mismatch

• TICON: Doodson 056554
• Neaps: Doodson 056555 (xdo [0, 5, 6, 5, 5, 5, 5])
• The 6th digit is 5 in Neaps but 4 in the TICON manual.

2. MKS2 — Wrong Doodson and speed

• TICON: Doodson 245555, speed 28.4350877
• Neaps: Doodson 257555, speed 29.0662415 (xdo [2, 5, 7, 5, 5, 5, 5])
• The JSON values for MKS2 actually correspond to what the TICON manual calls 3N2 (Doodson 257555, speed 29.0662415). There is no entry in Neaps with the correct MKS2 values.

3. 3N2 — Missing as a semidiurnal constituent

• TICON: Doodson 257555, speed 29.0662415
• Neaps: The alias "3N2" exists on N6 (a species-6 constituent with speed 85.3191886), which is an entirely different thing. There is no semidiurnal 3N2 in Neaps. The Doodson/speed slot that should belong to 3N2 is currently occupied by the misidentified MKS2.

4. 3L2 — Null Doodson + wrong speed

• TICON: Doodson 265555, speed 29.5331208
• Neaps: xdo null, speed 88.5854367, defined as compound [["L2", 3]]
• Neaps treats 3L2 as 3× the L2 frequency rather than as its own semidiurnal constituent with Doodson number 265555.

5. T3 — Null Doodson + wrong speed

• TICON: Doodson 381555, speed 44.9589333
• Neaps: xdo null, speed 44.9384
• Missing Doodson number. Speed is off by 0.0205 deg/hr.

6. R3 — Null Doodson + wrong speed

• TICON: Doodson 383555, speed 45.0410706
• Neaps: xdo null, speed 45.0616001
• Missing Doodson number. Speed is off by 0.0205 deg/hr.

---

This exposes a fundamental issue with the current Neaps implementation, which is that it assumes constituent definitions are standard and shared. Ultimately, I think it needs to support loading external definitions.

[bkeepers]

openwatersio/neaps#253 is open to correct the last 4 constituents, but two still remain:

1. Sa - Doodson 056555 (p'=0, Schureman convention) vs TICON's 056554 (p'=−1, Doodson/CTE convention). Both are legitimate; the difference is whether the solar perigee term is included.

2. MKS2 — TICON lists Doodson 245555, speed 28.4350877. IHO defines it as 257555, speed 29.0662415, which is the correct M2+K2−S2 compound. TICON's values are internally consistent (their speed matches their Doodson) but describe a different spectral line than what "MKS2" conventionally means. Whether TICON's analysis software actually extracted at 28.435 or 29.066 deg/hr is unknown. The practical impact is small — median MKS2 amplitude is 2.1 mm, though 8 stations (Bay of Fundy, Ireland, NYC, Netherlands) exceed 10 cm.