Updated Mar 29 2026 - An interactive browser with an enhanced view of the catalog is available at https: //eclipsedb. org/ This deposit contains two long-baseline eclipse catalogs computed from the JPL DE441 planetary ephemeris via the NASA NAIF SPICE toolkit, spanning 30, 000 years (year −13000 to +17000, astronomical year numbering): solarₑclipsefinal. txt — 70, 647 solar eclipses lunarₑclipsefinal. txt — 71, 914 lunar eclipses README. md — full column definitions, type codes, coordinate conventions, and usage notes FoundationsforStatisticalEclipseIdentificationInArchaeoastronomyURL. pdf — companion paper describing the three-tier validation framework and a statistical methodology for archaeoastronomical eclipse identification The catalogs extend the verified eclipse record by a factor of six relative to the NASA Five-Millennium Canon (Espenak & Meeus 2006/2009), which covers years −1999 to +3000. Each record includes eclipse type, gamma, magnitude, duration, tropical longitude, Lahiri sidereal longitude, Saros series, and ΔT. Validation: Tier 1 external audit against the NASA Canon yields 100% Saros agreement and 99. 0% type agreement across 11, 896 matched events. Tier 2 internal residuals show gamma mean = 0. 000, σ = 0. 0006. Tier 3 physical consistency checks recover the nodal precession period (9. 3064 yr, FAP ≪ 10⁻¹⁰) and the IAU precession rate (50. 30 arcsec/yr) from the catalog data, confirming orbital coherence across the full 30, 000-year span. Important: Geographic parameters (longitude of greatest eclipse) are ΔT-sensitive and carry growing uncertainty for pre-telescopic epochs — approximately ±62° of longitude at 2000 BCE and ±104° at 4000 BCE. Eclipse type, gamma, tropical longitude, duration, and Saros membership are ΔT-independent and are reliable across the full catalog span. See README and companion paper for details. Source code is not published. The methodology is fully described in the companion paper included in this deposit.
Anish Kumar (Sun,) studied this question.