Moonrise, moonset and moon phase

How these moonrise and moonset times are computed

The Moon's position comes from the truncated ELP-2000/82 series published by Jean Meeus in Astronomical Algorithms(chapter 47), which is stated to be accurate to about 10 arcseconds in ecliptic longitude and 4 arcseconds in latitude. Those coordinates are converted to right ascension and declination using the true obliquity of the date, then to your local sky through the apparent sidereal time at Greenwich. Because the Moon is close, the geocentric position is corrected for parallax with the rigorous reduction of chapter 40, including the flattening of the Earth: an observer on the surface sees the Moon up to about a degree lower than an observer at Earth's centre would.

Rise and set use the Moon's own horizon convention, not the Sun's. For the Sun this site uses an altitude of -0.833°, which combines refraction with the solar semidiameter. For the Moon the standard altitude is h0 = 0.7275 × π − 0.5667°, where π is the equatorial horizontal parallax. At the Moon's mean distance that works out to about +0.125°: the Moon's centre is geometrically above the horizon at the moment its upper limb appears, the opposite sign to the Sun, because the parallax correction is larger than the refraction and semidiameter together.

The crossing itself is found by search, not by a closed-form half-day arc. A solar formula assumes the declination barely changes across a day; the Moon's can move by more than six degrees. So the tool walks the real local day in five minute steps, brackets every sign change of altitude minus h0, and bisects each bracket to the second. A quadratic fitted through each consecutive triple of samples catches grazing events, where the Moon appears and disappears again within a single step. The day window itself is the genuine civil day in the chosen time zone, found by binary search on the browser's own IANA database, so a 23 h 00 m spring-forward day and a 25 h 00 m fall-back day are searched at their real length.

Illumination is the real Sun-Moon-Earth phase angle from Meeus chapter 48, not a cosine of the Moon's age; the two differ by up to about two percentage points near the quarters. The figure shown is geocentric, which is what almanacs publish. Phase instants (new, first quarter, full, last quarter) come from the chapter 49 series and match published tables to within a couple of minutes.

Phase namescome from one definition only: the Moon's elongation, the Sun-Earth-Moon angle. Almanacs put first quarter at exactly 90° of elongation and last quarter at 270°, and this site calls a phase by its name for 6.5° either side of those angles, about half a day. Naming phases by the Moon's age in days instead looks equivalent and is not: the orbit is elliptical, so age and elongation drift apart by up to roughly half a day, which is enough to hang the words "first quarter" on the wrong evening. The tool, the geometry diagram below it and the phase table all read that single definition.

Altitude above the horizon includes standard atmospheric refraction, which lifts a body at the horizon by about 34 arcminutes. Below the horizon there is nothing anyone can observe, so that correction is faded smoothly out between 1° and 4° below the horizon and is exactly zero deeper than that: an altitude quoted for a Moon well under the horizon is the plain geometric one. Fading rather than cutting matters because the figure is printed. A hard cut puts a step in it, and a reader dragging the time slider through a low Moon would watch the number jump.

What is not modelled

• Your horizon. Every time here assumes a flat horizon at sea level. A mountain to the east delays moonrise; standing on a hill brings it forward. Terrain can move a real moonrise by several minutes, and by far more in a valley.
• Weather. Refraction is the standard sea-level value. Real refraction varies with pressure and temperature, and unusual air near the horizon can shift a rise or set by a minute or two on its own.
• Altitude above sea level. Height lowers the horizon and makes the Moon rise earlier. That correction is not applied.
• Sub-arcsecond terms. The series is truncated, and Delta T beyond the published NASA polynomial range (1900 to 2150) is extrapolated, so dates far outside this century drift.

Checked against the US Naval Observatory's data services and NASA JPL Horizons: rise, set and transit agree to within about a minute at every reference point tested, including 64 to 65 degree latitude cases, and azimuth agrees with Horizons to about a thousandth of a degree. The regression suite lives in scripts/test/moon.test.ts.

Days with no moonrise, no moonset, or two moonrises

The Sun rises once a day almost everywhere. The Moon does not, and that surprises people who assume a missing row is a bug. Successive moonrises are about 24 h 50 m apart, because the Moon moves eastward against the stars while Earth turns. Each moonrise therefore lands roughly fifty minutes later than the last, and about once a month it steps clean over a calendar-day boundary. That date has no moonrise at all. The same thing happens to moonset, on a different date.

The reverse case is rarer and more fun: two moonrises inside one local day. It needs a moonrise just after local midnight, so the next one, 24 h 50 m later, still fits before the day ends. Reykjavik gets one on July 7, 2026, with moonrises at 00:04 and 23:55 and a single moonset in between. A clock change can also make room for it, because a fall-back day is 25 h 00 m long.

Far from the equator the Moon can also stay up or stay down for a whole day, exactly like the midnight sun and polar night but on the Moon's monthly rhythm rather than the Sun's yearly one. On January 15, 2026 the Moon never rises at all over Longyearbyen at 78° north. Every one of these cases is written out in words by the tool above, with the neighbouring rise or set named, so a blank is never left for the reader to interpret.

Why the Moon has phases

Half of the Moon is always in sunlight, exactly like Earth on the day and night globe. What changes is how much of that lit half faces us as the Moon orbits Earth each month. When the Moon sits between Earth and the Sun we see the dark half (new moon); two weeks later Earth sits in the middle and we see the whole lit half (full moon). The terminator line you can watch crossing the Moon is the same sunrise and sunset line this site draws across Earth.

That geometry is also why the Moon's rising time tracks its phase. A full moon is opposite the Sun, so it rises near sunset and sets near sunrise. A new moon is beside the Sun, rises near sunrise, and is lost in daylight. A first quarter moon rises around midday and is highest around sunset. If the tool above says a bright gibbous moon rises in the late evening, that is the geometry, not an error.