Should you plan August 12 around the 100 meteors an hour every forecast promises, or around the 25 that NASA’s own fact panel quotes?
Welcome to FreeAstroScience. We opened NASA’s Perseids page, the American Meteor Society calendar and the International Meteor Organization’s 2026 entry side by side this morning. Between them they quote five different rates, so this piece works out which one belongs in your plans.
The 2026 Perseid meteor shower peaks overnight from August 12 into August 13, and a realistic count from a dark rural site is 25 to 75 meteors an hour, not the 100 usually quoted. Zenithal hourly rate is the technical name for that 100, and the standard assumes a perfectly dark sky with the shower’s origin point directly overhead. Neither condition holds anywhere on Earth that night.
NASA’s Perseids page cannot decide. Its main text says about 50 to 100 meteors are seen per hour under ideal dark skies. Further down the same page, a fast-facts panel says about 25 meteors per hour. One page, one shower, and a factor of four between the two numbers a reader is most likely to write down and take outside.
Three official pages, five different numbers
Open the other two and the gap widens. The American Meteor Society’s shower calendar puts the Perseids at 50 to 75 shower members per hour from rural locations at maximum. Over at the International Meteor Organization’s 2026 calendar, the shower carries a ZHR of 100 and runs from July 17 to August 24, peaking on the night of August 12 into 13. That same page then puts its own rural figure at 30 to 50 an hour, which is not the AMS number for the same kind of sky. Both windows match the graphic at the top of this page exactly, which is a relief. The rates match nothing.
One detail deserves flagging before you trust any of it. When we opened the AMS calendar on August 6, 2026, its Perseid entry still carried last year’s dates throughout. Activity ran July 14 to September 1, 2025, and the Moon note described where the 2025 Moon would set. So the 50 to 75 figure describes the shower’s general character rather than a 2026 forecast. Reading it as this year’s prediction would be reading it wrong.
Table 1 — Perseid rate figures as published, and the conditions each one assumes (sources as listed, retrieved 6 August 2026)
| Source | Rate quoted | Conditions assumed |
|---|---|---|
| NASA, page text | 50 to 100 per hour | ideal dark skies |
| NASA, fast facts | about 25 per hour | dark skies |
| American Meteor Society | 50 to 75 per hour | rural location, at maximum |
| International Meteor Organization | ZHR 100 | radiant overhead, limiting magnitude 6.5 |
| International Meteor Organization | 30 to 50 per hour | rural location, at maximum |
Only the last row states its conditions in a way you can check against your own back garden. Conditions turn out to matter more than the number sitting in front of them.
What does a zenithal hourly rate actually measure?
Definitions settle this one, and the International Meteor Organization’s glossary gives one without hedging. A zenithal hourly rate is “the number of shower meteors per hour one observer would see if his limiting magnitude is 6.5mag and the radiant is in his zenith”. Two pieces of jargon carry that sentence. Radiant names the point on the sky the meteors appear to stream away from, which for this shower sits in Perseus. Limiting magnitude names the faintest star a person can still pick out, and 6.5 describes a sky with no Moon, no cloud, no haze and no town anywhere near it.
Both conditions are fictions.
They are useful fictions, and forecasts routinely leave that part out. A ZHR exists so a count made from a Spanish hillside can be set against a count made from a Finnish lakeside without either observer arguing about whose sky was better. Nobody ever meant it as a promise about what you will see. Treating it as one is like reading a car’s laboratory fuel figure and expecting it in city traffic.
Where the radiant actually sits above Rimini
FreeAstroScience publishes from Rimini, on the Adriatic coast at latitude 44.06° north, so that is the back garden we ran the geometry for. Perseus puts the radiant at right ascension 03h 13m, declination +58°, a position the AMS calendar lists beside the rate. Feed those two numbers and the date into spherical trigonometry and the night takes a clear shape.
At 2 a.m. on August 13, the radiant stands roughly 44° above the horizon from here. By 4 a.m. it is close to 60°. Its ceiling, in the last hour before the sky greys out, is roughly 76°, because 76 is all that Rimini’s latitude and the radiant’s declination together allow. Overhead is not on the menu. From London or Vancouver the radiant climbs closer to the zenith, and from Rome or New York it stays lower.
Geometry like that is not a technicality. A radiant sitting 44° up throws its meteors across a foreshortened patch of sky, and a good fraction of the shower is happening below your horizon with the Earth in the way. Watch at 11 p.m. because you want an early night, and you are watching the worst hour of the display. Every hour you wait past midnight buys altitude, and altitude buys meteors. That is exactly why the graphic’s advice to stay out until dawn is the most valuable line on it.
Turning that altitude into an exact count needs one more quantity: the population index, a number describing how a stream’s meteors are shared out between faint ones and bright ones. None of the four pages we read published a population index for the 2026 Perseids. We are not going to invent one. So the honest output of this section is a direction rather than a figure: watching early costs you meteors, and the cost shrinks steadily from midnight to dawn.
Why is your sky taking meteors away from you?
Darkness is the second fiction, and most of the missing meteors go here. Limiting magnitude 6.5 describes a sky almost nobody reading this has standing access to. A typical suburban garden falls a long way short of that, and sky glow deletes the faintest meteors before it touches the bright ones. Faint meteors dominate the stream. NASA puts Perseid fireballs down to larger particles of cometary material, which is the same point from the other end: the ones people remember afterwards are the exception, not the bulk.
Three things eat your count, and only one of them sits outside your control.
- Sky glow, from the town you did not drive far enough to escape.
- Your own eyes, which need about 20 minutes to reach full dark adaptation and surrender it to two seconds of phone screen.
- Cloud.
Both of the first two repay real effort. Driving 40 minutes to somewhere genuinely dark will do more for your night than any equipment purchase. And the 20 minutes our graphic asks you to wait is the instruction people skip most often, usually by checking a phone. Something slower is arriving overhead, which we covered in our report on satellite mega-constellations and the future of the night sky. Worse, the sky you drive out to find is itself getting brighter, and the trend does not favor the observer.
The comet running the wrong way
One factor works in your favor, and it explains why the Perseids stay worth the drive when fainter showers are not. Debris for this shower comes from comet 109P/Swift-Tuttle, a body 26 kilometers across. JPL’s Small-Body Database gives its orbit an inclination of 113.45°. Any inclination past 90° means the comet circles the Sun backwards, against the traffic, so Earth meets its dust close to head-on instead of catching it from behind.
That collision geometry is why a Perseid arrives at 59 kilometers per second, the 133,200 miles per hour printed on the graphic. Speed is what makes them bright. A fragment no larger than a grain of sand squeezes the air ahead of it hard enough to ionize it, and the IMO glossary ties the intensity of the resulting glow straight to entry velocity. What you see is that lit air, not the grain, which is gone in well under a second. Bright meteors punch through mediocre skies. Faint ones do not, which is why the Perseids survive a suburban horizon better than most showers on the calendar.
The same database puts Swift-Tuttle’s orbital period at 48,681 days, or 133.3 years, and carries it out to 51.2 times the Earth’s distance from the Sun before it turns around. Compare the Orionids, whose parent is Halley’s Comet on a 76-year loop. We walked through that shower on its own terms in our October guide to the Orionids. A different comet, and a separate night of planning this piece does not try to cover.
The Moon is the one thing going right this year
One variable lands in your favor, and it is settled by an event that has nothing to do with meteors. On August 12, 2026 a total solar eclipse sweeps Greenland, Iceland, Spain, Russia and a small part of Portugal, according to NASA’s eclipse listing. That only happens at new Moon. So the Moon is new on the afternoon of the peak day. By the time the radiant is climbing after midnight, it is a sliver under a day old that set with the Sun hours before.
That adds up to no Moon at all, which is the best sky the calendar can hand a meteor shower. Compare 2025, when the AMS entry called a half-lit Moon setting near 1 a.m. a favorable coincidence, because it cleared the sky just as the good hours began. Nothing needs clearing this year. Anyone also watching the eclipse that day should read our guide to choosing eclipse glasses that are genuinely certified first, because that half of August 12 can injure you. The meteors after midnight cannot.
What can no forecast tell you?
Three gaps sit inside everything above, and they belong in your expectations too. First is the population index we could not source, which means the numbers here set a direction and a plausible band rather than a prediction anyone should be held to. Second, we are leaving out the dust-trail modeling that tries to flag unusually strong years, because none of the four pages we read carried a 2026 prediction of that kind. That work deserves its own article. Waiting for it would not change what you should do on the night of August 12 anyway.
Third is weather. Weather outranks every other factor in this piece combined, and a clear suburban sky beats a clouded dark-sky reserve by a margin no arithmetic will close. No meteor forecast published anywhere knows what your Wednesday looks like.
One more limit deserves stating plainly. Run the same arithmetic for 30° south and the radiant tops out near 2°, which is the horizon rather than the sky. Readers there should treat every number in this piece as somebody else’s forecast.
The number to take outside on Wednesday night
Plan for 25 to 75 meteors an hour from a genuinely dark site after midnight, and treat anything above that as generosity. That 100 is a ZHR, defined for a radiant directly overhead under a limiting magnitude of 6.5, and from Rimini the radiant tops out near 76° and never gets there. The Moon, at least, is entirely on your side, because the total solar eclipse of August 12 fixes it as new.
We wrote the arithmetic out in full because a number handed over without its conditions is not information but decoration, and you deserve better than decoration on the one clear night of your summer. A mind that stops checking gets fed anything. Argue with us: go outside on Wednesday and hold your own count against the table above. Once observed rates for 2026 are published, come back and we will run them against these five figures. Gerd Dani, FreeAstroScience, Rimini.
Sources
- NASA Science. Perseids. Published 20 November 2017, last updated 3 April 2025. https://science.nasa.gov/solar-system/meteors-meteorites/perseids/
- International Meteor Organization. Meteor Shower Calendar, 2026 Perseid entry. Retrieved 6 August 2026. https://www.imo.net/resources/calendar/
- International Meteor Organization. Glossary, definition of zenithal hourly rate. Retrieved 6 August 2026. https://www.imo.net/resources/glossary/
- American Meteor Society. Meteor Shower Calendar 2026-2027, Perseid entry. Retrieved 6 August 2026. https://www.amsmeteors.org/meteor-showers/meteor-shower-calendar/
- Jet Propulsion Laboratory, NASA. Small-Body Database Lookup: 109P/Swift-Tuttle. Orbital elements and physical parameters retrieved 6 August 2026. https://ssd.jpl.nasa.gov/tools/sbdb_lookup.html#/?sstr=109P
- NASA Science. Future Eclipses, total solar eclipse of 12 August 2026. Retrieved 6 August 2026. https://science.nasa.gov/eclipses/future-eclipses/
- FreeAstroScience and MattyAtoms. The Best Meteor Shower: The Perseids, infographic, 2026. Activity window, peak dates, particle size, entry speed and viewing guidance.
