Banshoreki萬象暦

Banshoreki › Verification

Verification

Every astrology site says its numbers are accurate. This page shows the measurement. The figures below are read from the same data files as the tables, so they cannot drift from the tables.

Western: against NASA JPL Horizons

Result
Sun and planets: max 0.61″ · asteroids: max 0.66″ · sidereal time: max 0.71 s
100 dates · built 2026-09-05

100 dates drawn with a fixed random seed between 1900 and 2100. For each, the apparent geocentric ecliptic longitude of the Sun, Moon, planets and the asteroids Chiron, Ceres, Pallas, Juno and Vesta is compared with Horizons (ephemeris DE441), and local sidereal time is compared as well. The Moon comes from Astronomy Engine rather than a Horizons table, so its difference grows after 2050 because of the ΔT extrapolation; that row is shown in red rather than hidden. Sidereal time after 1962 differs by the UT1−UTC term that Horizons includes (under one second).

Open the full table (100 rows, every body)

Eastern: against the National Astronomical Observatory of Japan

Result
Solar terms: 576 compared, max 0.53 min · new moons: 297 compared, max 0.62 min · day mismatches: 0
2004–2027 · built 2026-09-05

The 24 solar terms and every new moon from 2004 to 2027, as published in NAOJ's annual calendar bulletin (to the minute), are compared with the calculation. Solar terms decide the month pillar in Four Pillars astrology, so a difference across a day boundary would change a chart; the day mismatches are counted separately.

Open the full table (solar terms and new moons)

Other checks (not on this page, recorded in the repository)

Sources: NASA JPL Horizons (ssd.jpl.nasa.gov), NAOJ calendar bulletin (© National Astronomical Observatory of Japan), Astronomy Engine, Swiss Ephemeris (used only as a yardstick, not distributed).