A lunar eclipse occurs when the Sun, Earth, and Moon align closely enough for the Moon to pass through Earth's shadow, and it can only happen at full moon, on dates that fall within roughly 17 to 38 days of a node crossing. Because the Moon's orbit tilts about 5.1 degrees to Earth's orbital plane, most full moons miss the shadow entirely. That single geometry fact explains why eclipses arrive in seasons, why they cluster in pairs or trios, and why roughly two lunar eclipses occur each year but not every one is visible from where you stand.
This explainer goes deeper than a calendar entry. It covers the alignment mechanism, the umbra and penumbra geometry, the timing rules that let you predict an eclipse years out, the edge cases that decide whether you see totality or a partial shading, and the practical steps for observing one safely.
The alignment mechanism: nodes, syzygy, and why full moon is not enough
A lunar eclipse needs three bodies on a near-line, a configuration astronomers call syzygy. The Sun and Earth define a line, and Earth's shadow extends away from the Sun along that line. For the Moon to enter the shadow, it must be close to that line at the moment it reaches full phase.
The obstacle is orbital inclination. The Moon's orbit is tilted about 5.1 degrees relative to the plane of Earth's orbit, the ecliptic. The two points where the Moon's path crosses the ecliptic are the nodes. Only near a node can the Moon's center pass close enough to the shadow axis to be eclipsed.
Here is the mechanism one level deeper. Earth's shadow is a cone that narrows with distance. At the Moon's average distance of about 384,400 kilometers (238,900 miles), the umbra — the fully shadowed core — spans roughly 9,200 kilometers (5,700 miles) in diameter. The Moon is about 3,475 kilometers (2,160 miles) across. The cone is nearly three lunar diameters wide at that distance, which is why the Moon can be fully engulfed without the alignment being perfect. The shadow is forgiving of small errors, but not of 5-degree errors.
The nodes themselves drift westward around the ecliptic, completing a circuit in about 18.6 years. That regression is the clock behind eclipse seasons.
Quantified context: eclipse seasons, frequencies, and durations
An eclipse season is the window during which the Sun is close enough to a node that a full moon can graze the shadow. Each season lasts roughly 31 to 38 days, and seasons recur about every 173 days — a little under six months.
Within a season you usually get one lunar eclipse, sometimes two, rarely three. Because the Sun must be near a node, and a node alignment can catch two consecutive full moons at the edges of the window.
| Eclipse type | What the Moon enters | Typical frequency | Rough share of lunar eclipses |
|---|---|---|---|
| Penumbral | Outer shadow only | 1–2 per year | About 35 percent |
| Partial | Umbra, not fully | Roughly 1 per year | About 30 percent |
| Total | Entirely in umbra | Every 1–3 years | About 35 percent |
The numbers shift year to year because of the 173-day season spacing against the 365-day calendar. Some calendar years contain four lunar eclipses across two seasons. Others contain two. The long-run average is about 2.4 lunar eclipses per year, of which roughly 0.8 are total.
Duration scales with how centrally the Moon crosses the shadow. A grazing penumbral eclipse may last under an hour of visible shading. A deep total eclipse can run about 5 to 6 hours from first penumbral contact to last, with totality itself lasting up to roughly 107 minutes. The maximum totality duration is capped by the Moon's speed through the shadow, about 1 kilometer per second relative to the shadow axis, and by the shadow's width.
Plain language: the shadow is wide, the Moon moves fast, and the difference between a brief total eclipse and a long one is how close to the shadow's centerline the Moon travels.
The eclipse timeline: what happens and when
Every lunar eclipse follows the same contact sequence. Knowing the labels lets you plan when to look up.
| Phase | What happens | Typical duration into the event |
|---|---|---|
| P1 — penumbral start | Moon touches outer shadow | 0 minutes |
| U1 — partial start | Moon touches umbra; first dark bite | About 60–90 minutes |
| U2 — totality begins | Moon fully inside umbra | About 90–150 minutes |
| Greatest eclipse | Moon closest to shadow axis | Midpoint of totality |
| U3 — totality ends | Moon begins leaving umbra | About 90–150 minutes after U2 |
| U4 — partial ends | Moon fully out of umbra | Shortly after U3 |
| P4 — penumbral end | Moon leaves outer shadow | End of event |
The penumbral phases are subtle. The shadow's outer edge is a gradient, not a line, because the Sun is a disk rather than a point source. During P1 to U1, a careful observer notices a dusky shading on one limb of the Moon, but casual viewers often see nothing until the umbra arrives.
That is the counterintuitive detail worth flagging: the "start time" printed in many calendars is P1, yet the visible show usually begins at U1. If you arrive at P1 expecting drama, you will conclude the eclipse was a dud. Arrive at U1.
Edge cases: why some eclipses look different
Central versus non-central crossings. A central eclipse takes the Moon near the shadow axis and produces long totality. A non-central eclipse clips the shadow, producing brief totality or a partial event. Same label, very different experience.
The Danjon effect and atmospheric color. Totality is not black. Sunlight refracts through Earth's atmosphere and bends into the umbra, reddening as it goes. The brightness of the red Moon depends on how much dust and aerosol sits in the stratosphere. After major volcanic eruptions, total eclipses can turn nearly invisible, dark gray-brown rather than copper. That is why the Danjon scale, a 0-to-4 brightness rating used by observers, exists. A high value means a bright copper Moon, a low one means a dark, hard-to-see disk.
The distance variable. The Moon's distance varies from about 356,500 to 406,700 kilometers (221,500 to 252,700 miles) between perigee and apogee. A closer Moon appears larger and its disk subtends more of the shadow, marginally changing duration and the visual scale of totality.
The "blood moon" label. Any total lunar eclipse can look red. The term is marketing, not a distinct astronomical category. Every total lunar eclipse is a blood moon in the same sense.
Penumbral eclipses at the threshold. Some penumbral eclipses are so shallow that the shading is essentially unobservable without photometry. Others, where the Moon passes deep into the penumbra, show an obvious darkening on one side. Two penumbral eclipses on a calendar can be entirely different events.
Misconceptions: solar versus lunar, safety, and the full-moon trap
Solar and lunar eclipses are mirror images. A lunar eclipse happens at full moon, when Earth sits between Sun and Moon. A solar eclipse happens at new moon, when the Moon sits between Sun and Earth. The confusion is common because both are called eclipses.
Lunar eclipses are safe to view without filters. Unlike solar eclipses, where direct viewing can injure the retina, a lunar eclipse is just the Moon in shadow. You can watch it with the naked eye, binoculars, or a telescope with no protective filter. This is the most common practical misconception worth correcting.
Full moon does not guarantee an eclipse. The Moon is full roughly every 29.5 days, but the shadow alignment is only right during a node window. Most months, the full moon sails above or below the shadow cone. If full moon always meant eclipse, we would see about 12 lunar eclipses a year. We see roughly two.
Eclipses do not harm pregnancy or food, despite persistent folk claims in some regions. There is no physical mechanism by which a lunar eclipse affects a pregnancy or spoils food. The shadow is a geometric event, no different in physical effect from the Moon passing behind a hill.
The Moon does not "go out." It is lit by refracted sunlight through Earth's atmosphere, not by direct sunlight, which is why it reddens rather than vanishing.
Red-flag warnings for planning:
Do not assume a printed eclipse time is your local time; convert from UTC carefully, including daylight saving shifts. Do not trust a generic "visible worldwide" claim; roughly half of lunar eclipses are not visible from any given location. Do not plan on a penumbral eclipse being visually dramatic; the shading can be imperceptible. Do not use solar-eclipse filters on a lunar eclipse; they are unnecessary and can dim the view to nothing.
Practical actions: how to catch and time the next eclipse
Step 1: Identify the phase you want. Decide whether you want totality, a partial bite, or the faint penumbral shading. For most people, U1 to U4 is the meaningful window. If totality is the goal, you need only U2 to U3.
Step 2: Convert UTC to local time. Published eclipse tables give contact times in UTC. Add or subtract your time zone offset, then apply daylight saving if active. A two-minute conversion error is common; a full hour error is more common.
Step 3: Check the visibility map for your region. An eclipse can be total but occur while the Moon is below your horizon. The Moon must be above the horizon during the phase you want. If the eclipse peaks after moonset, you will see only the early phases, or nothing.
Step 4: Find a dark site with an unobstructed view. Lunar eclipses are visible from cities, but darker skies make the red color and the penumbral gradient easier to see. The Moon does not need dark adaptation the way faint stars do.
Step 5: Set two alarms. One for U1, one for U2. The gap between them is your buffer for cloud cover and setup.
Thresholds that matter: totality can last from seconds to about 107 minutes. Partial phases can run 1 to 3.5 hours. The entire event from P1 to P4 rarely exceeds 6 hours. If a published totality duration is under 10 minutes, expect a brief, shallow event with a bright rim of the Moon still catching sunlight.
If you want to predict eclipses yourself, the Saros cycle is the tool. A Saros is about 18 years and 11 days, and it produces a near-repeat of eclipse geometry. A Saros series produces eclipses of the same general character every cycle. Professional caveat: Saros repeats are approximate, and the visibility geography shifts with each return.
Limitations: what timing predictions cannot tell you
Eclipse timing is now highly precise for contact times, typically to within seconds, because the lunar orbit is well modeled. What remains less predictable is the appearance. The Danjon brightness of totality depends on stratospheric aerosol loading, which changes after volcanic eruptions and large wildfires. You can know the exact minute of totality and still be surprised by whether the Moon is copper-bright or nearly black.
Local weather is the bigger limit. No eclipse prediction can guarantee clear skies. The visibility map tells you the Moon's position relative to the horizon, not the cloud cover.
The distinction between "when" and "where" is the practical bottom line. The question when is the lunar eclipse has a precise answer in UTC for any given event. Whether you personally see it depends on two variables the tables cannot supply: your horizon and your sky.
For observers planning ahead, the reliable approach is to treat the published contact times as fixed and your location and weather as the uncertainties to manage. That is what separates a satisfying eclipse night from a missed one.