Moonrise and moonset for sixteen cities, on any date
Moonrise and moonset for sixteen cities on any date, including the days with no rise at all — which happens far more often than most tables admit.
Local clock times. A day with no rise or no set is not an error — the Moon rises on a schedule the Sun does not share, and near the poles it can skip a date entirely.
| City | Moonrise | Moonset |
|---|---|---|
| New York · United States | 00:26 | 14:02 |
| Chicago · United States | 00:21 | 13:59 |
| Denver · United States | 00:37 | 14:07 |
| Los Angeles · United States | 00:40 | 13:56 |
| Honolulu · United States | 01:39 | 14:29 |
| Anchorage · United States | 00:59 | 15:39 |
| Mexico City · Mexico | 01:36 | 14:27 |
| São Paulo · Brazil | 01:38 | 13:22 |
| London · United Kingdom | none | 14:18 |
| Reykjavík · Iceland | 00:36 | 16:17 |
| Tromsø · Norway | none | 15:02 |
| Cape Town · South Africa | 02:25 | 13:40 |
| Nairobi · Kenya | 01:32 | 13:53 |
| Singapore · Singapore | 01:52 | 14:18 |
| Sydney · Australia | 02:22 | 13:25 |
| Longyearbyen · Svalbard, Norway | 22:58 | 16:54 |
Times are for the center of the disc at an altitude of +0.1°, which is how the Moon is conventionally defined as risen: its parallax lowers it more than refraction lifts it, so it is already visible when its center is still below the geometric horizon. "No set" means the Moon stayed above the horizon for the whole local day.
A moonrise time is the answer to a question the Sun never asks: the Moon is not tied to the solar day. It moves about 13° a day eastward against the stars, so it comes up roughly 50 minutes later each night — and that average hides a range from roughly 20 minutes to over an hour, depending on the angle the ecliptic makes with the horizon at your latitude.
The table above gives rise and set for sixteen cities in their own clock time. Missing entries are real.
Why a date can have no moonrise
A lunar day — one full turn relative to the Sun, which is what governs when the Moon is up — is about 24 hours 50 minutes. A calendar day is 24 hours. So the Moon’s schedule slides out of step by about an hour a day, and roughly once a month the rise lands after midnight on one date and the next rise lands before midnight two dates later, leaving a date with none.
That is not an edge case to be patched with interpolation. It is a direct consequence of 29.53 days of lunation not dividing 24 hours, and a table that shows a rise every day is wrong on roughly 12 dates a year.
The horizon convention
For the Sun, refraction dominates: the Sun is 34′ higher than its geometry says. For the Moon, parallax
works the other way, and it is larger. The Moon is close enough that an observer on the surface sees it about
0.95° below where a geocentric observer would — more than the refraction that lifts it. The standard
convention (Meeus, chapter 15) folds both into one number: the Moon counts as risen when its center
reaches an altitude of 0.7275·π − 34′, which at a typical distance is about +0.1°.
It is a small number with a visible consequence: the Moon is above the horizon a couple of minutes longer than a naive geometric calculation predicts, on every rise and every set.
Why the solver samples instead of solving
For the Sun, the hour-angle formula is enough, because the Sun’s declination changes little in a day. For the Moon it changes by up to 5° in a day, and at high latitude that is the difference between a crossing and a grazing pass. A closed-form solution built on a constant declination can be two hours out exactly where it matters most.
So src/lib/sky.mjs samples the altitude every ten minutes over the local day and bisects each sign change.
Against the USNO’s tables for New York, Sydney, Tromsø, Reykjavík and Svalbard the worst disagreement is
0.5 minutes, and one of those cases — Tromsø, November 3 — is a day with a set and no rise, which is the
case a symmetry-assuming solver gets wrong.
If you want the underlying mechanism rather than a lookup, the fifty-minutes-later page explains where the drift comes from, and the lunar standstill page explains why the range is not constant across the 18.6-year cycle.
Computed for a standard horizon at a city center. Terrain and elevation are not modelled.
Frequently asked questions
- Why is there no moonrise on some dates?
- Because the Moon gains roughly fifty minutes a day on the Sun. A lunar day is about 24 hours 50 minutes long, so a rise that happens at 23:40 one evening happens at 00:30 the next morning, and the calendar day in between has no rise in it. Most months have one or two such dates, and a table that always shows a rise is filling the gap with a guess.
- Why does the Moon sometimes set in the morning and rise in the evening?
- The Moon is up at whatever time its position in the sky says, which drifts by about 50 minutes a day against the Sun's schedule. Around full moon it rises at sunset and sets at sunrise; around new moon it is up with the Sun and invisible. That drift is the same fact as the missing-rise dates, seen from the other side.
- Why do these times differ from a smartphone weather app?
- Usually by less than a minute, and the difference is the horizon convention. This table uses the standard one for the Moon — the center of the disc at +0.1° altitude, where parallax lowers the Moon almost exactly as much as refraction lifts it. An app that treats the Moon like a star, or that uses the upper limb, will differ by one to three minutes.
- How is the polar case handled?
- Inside the polar circles the Moon can stay above or below the horizon for days at a time. The table prints "no set" or "no rise" instead of a time, because there is no crossing to report. Longyearbyen is in the list so that this is visible rather than theoretical.
Sources
- U.S. Naval Observatory, Astronomical Applications Department — the rise/set tables this site's lunar solver is measured against, agreement within 0.5 minutes across New York, Sydney, Tromsø, Reykjavík and Svalbard test cases (aa.usno.navy.mil/data/RS_OneDay)
- Meeus, Astronomical Algorithms, 2nd edition, chapter 15 — the rise/set definition for a body with a large parallax: h₀ = 0.7275·π − 34′, and the iterative treatment of a declination that changes during the day
- JPL Horizons (DE441) — cross-check for the Moon's right ascension, declination and distance that the altitude computation depends on (ssd.jpl.nasa.gov/horizons/)
- NASA Science — Moon: the 5.1° inclination of the lunar orbit to the ecliptic, and why the Moon's declination range is wider than the Sun's