Astronomers are still searching for answers behind this year’s unusual wave of loud and fiery meteor sightings. Over 3,000 people witnessed a slowly disintegrating daytime fireball over Western Europe. Hundreds more reported the sight—and sonic boom—of a 7-ton, 6-foot (2-meter) asteroid screeching above Ohio. March alone has already seen over 40 meteor cases, with yet another ripping through the sky over Texas last Saturday, breaking the sound barrier, before a fragment crashed into a north Houston home and ricocheted around one bedroom like a pinball.
Now, a new analysis published by the American Meteor Society (AMS) on Wednesday has confirmed just how much of a statistical outlier this 2026 barrage has been—as well as early indications of where all these rocks in our solar system might have come from.
“After years of stable baseline activity, something appears to have shifted,” according to AMS researcher Mike Hankey, who manages the society’s fireball reporting tools. “The signal is consistent across multiple metrics.”
According to those metrics—including total witness figures, the number of cases involving sonic booms, and the duration of the sightings—Hankey said, “Fireball activity has increased.”
Fireballs from outer space, loud enough to produce a sonic boom and witnessed by 50 or more people, have blitzed a trail through Earth’s atmosphere approximately once every three days since this year began, based on reports to the AMS.
“What makes 2026 unique is the combination,” Hankey wrote. “Prior high-sound years like 2021 and 2023 had elevated percentages but moderate event counts. In 2026, both the rate and the absolute count are high.”
Looking at meteor events with the highest number of witnesses—meaning 50 reports or more—30 out of 38 were meteors that were big, tough, and fast enough to produce a sonic boom (79%), which already makes the first quarter of 2026 an outlier historically. But Hankey also determined that the total number of mass sighting events and the volume of those witness reports were outliers, too. Excluding the phenomenal March 8, 2026 case over Western Europe, in which a whopping 3,229 people all reported the same fireball, the remaining 41 episodes so far this March still averaged about 67 witnesses per meteor, “more than double the historical norm,” Hanky noted.
In other words, while the total number of meteor cases has not deviated from researchers’ statistical expectations, the percentage of loud and well-documented cases did.
“Almost half of all March 2026 events with 10+ reports were seen by 50 or more people,” according to Hankey. “Events that would normally draw 25 [to] 49 witnesses instead drew 50, 100, or even 200+ witnesses. The distribution didn’t broaden—it shifted upward.”
Hankey cautioned that the AMS data for 2026’s meteor bombardment can only help develop witness-based trajectory estimates, not the more precise trajectories based on instrument data. But the sheer volume of witnesses does help us learn a bit about where these rocks came from.
Activity from a region of space known as the “Anthelion sporadic source,” defined as objects that hit Earth on their way deeper into our solar system toward the Sun, roughly doubled in 2026. A total of 12 meteors traced back to this Anthelion slice of the sky in 2026, with nearly 10 of those events apparently emanating from a single 1,000 square-degree patch.
Several of the biggest meteor events this month were traced back to this Anthelion region—including a March 9 fireball spotted by 282 people across the U.S. eastern seaboard and two fireballs that were reported 381 times over France across the following two days.
For now, Hankey believes that this current data can rule out a few hypotheses for what’s causing this uptick in meteors, or at least meteor sightings.
First, the Anthelion trajectories indicate that there’s no new cluster of asteroids entering Earth’s transit around the Sun—the sort of drifting space rocks that produce predictable annual meteor showers, like the Perseids every August.
Second, early material analyses of the fragments recovered in Ohio and Germany have had the mineral makeup of achondritic HEDs, one of the most common categories of meteorites on record. Hankey concluded that, for these reasons, it’s highly unlikely that any of these fireballs were crashing extraterrestrial spacecraft: “There is no evidence of anomalous trajectory behavior, controlled flight or non-natural composition,” he wrote in the AMS report. (Although, who’s to say aliens wouldn’t want to throw rocks at Earth.)
Hankey speculated that AI-chatbot advice might have helped more people report their sightings to AMS (one potentially very mundane explanation for the volume of reports), but there’s more than enough mystery left to warrant “serious investigation,” in his opinion.
“Whether this represents normal statistical variance,” he said, “an uncharacterized debris population, or something else entirely will require continued monitoring and further analysis.”
(Score: 3, Interesting) by khallow on Wednesday April 08, @01:36AM
How big is this area? There are 41,253 square degrees in a sphere. So ~1000 square degrees is roughly 2.4% of the sky, including the part we can't see because the Earth itself is in the way. If you look at visible sky with no obstructions to the horizon, it's about 5% of that.
In comparison, the Moon and Sun crudely are in the neighborhood of 0.2 square degrees. Which is 20,000 times less area. 1000 square degrees as a circle in the sky would be roughly 140 times wider in diameter than the Moon or Sun.
(Score: 4, Interesting) by JamesWebb on Wednesday April 08, @04:45AM (3 children)
Two populations, not one surge. 22 of the 67 Q1 fireballs track the anti-solar point across the sky — low-inclination, prograde orbits near 1 AU, shallow entry angles. Textbook anthelion source, but at 3-4x normal rate. Separately, 12 events come from high declination (>70°), steeply inclined orbits spread across all RA values. Different delivery path, same timing. Two HED meteorite falls 9 days apart — an Ohio eucrite and a NZ diogenite — are different rock types from the same mineral family. One disrupted parent body, fragments split into two orbital populations by resonance with Jupiter.
Quiet magnetosphere = louder fireballs. Fireballs arriving after 2+ quiet days averaged 287 witness reports vs. 79 during active periods — 3.66x. The March 8 event hit after 2 quiet days and drew 3,229 reports. The rocks aren't bigger. The detector (us) is more sensitive when the geomagnetic background is calm.
13-day periodicity. Daily event counts autocorrelate at lag=13 days, matching the Carrington half-rotation — the interval between solar wind sector boundary crossings. The magnetosphere's sensitivity cycles, and the observable fireball rate cycles with it. This periodicity only appears in high-flux years (2021 and 2026), because you need enough debris to sample the cycle.
Something broke apart, and we're flying through the wreckage on a predictable schedule.
(Score: 4, Interesting) by JamesWebb on Wednesday April 08, @10:30AM (2 children)
Kp index dropped to 0.33 today. Fireball witness counts should spike within 48 hours if the cochlea model holds.
(Score: -1, Redundant) by Anonymous Coward on Thursday April 09, @11:16AM (1 child)
Claude says:
(Score: 2, Interesting) by JamesWebb on Friday April 10, @12:15AM
You asked Claude to debunk it. Claude got it wrong. Let me show you why.
"No physical basis" for Kp/witness correlation.
Nobody claimed the magnetosphere affects human eyeballs. The correlation is in the data: across all 67 Q1 2026 events with 25+ witness reports, fireballs arriving after 2+ quiet days averaged 3.66x the witness count of those during active periods. That's not one cherry-picked event — it's a systematic pattern across the full dataset. The mechanism doesn't have to be "visual acuity." Ionospheric conditions, atmospheric transparency, and meteor breakup altitude all vary with geomagnetic activity. Claude refuted a claim nobody made and declared victory.
"Entirely speculative, no published support" for Carrington periodicity.
I downloaded 735,065 meteor orbits from the IAU Meteor Data Center's Global Meteor Network archive — an entirely independent instrument network, video-based, global coverage, 2021-2024. Daily rate autocorrelation:
2022 (54,046 orbits): AC at lag-27 = +0.60
2023 (58,884 orbits): AC at lag-27 = +0.39
2024 (110,889 orbits): AC at lag-27 = +0.40
That's the full 27-day Carrington rotation, confirmed across three years in a quarter-million events from a completely separate network. The original AMS lag-13 signal was aliased from this — 67 events is too sparse to distinguish lag-13 from lag-27. The underlying periodicity is real, it's the full solar rotation period, and it's stronger than the original report suggested.
Claude said "solar wind sector boundary crossings do not have a well-established connection to meteoroid flux." Correct — and nobody claimed they affect the flux. The rocks arrive regardless. The claim is that the observable rate is modulated. Those are different things. The GMN data — 250,000 events from cameras that don't care about Kp — still shows the 27-day signal, which means it's likely in the atmospheric entry physics, not just witness behavior.
"Post-hoc fitting / cherry-picking."
Claude's objection was that the signal only appears in high-activity years. That's not cherry-picking — it's the Nyquist theorem. You cannot detect a 27-day periodic signal with 3 events per month. You need enough events to sample the cycle. GMN confirms this: the signal is present in every year with sufficient coverage. 2021 GMN had only 25K orbits (small network) — weak signal. 2022-2024 scaled to 54-110K — strong signal. This is exactly what sampling theory predicts.
"Cochlea" was a typo from an earlier draft — should have been "correlation model." Apologies for the confusion.
The parent body question.
Claude says the two HED falls are "unrelated" because they have 98° angular separation at observation. That's not how orbital mechanics works. Fragments from a single disruption evolve into different orbital families over time — that's the entire basis of asteroid family identification (Hirayama families, Nesvorny clusters). Two HED achondrites falling 9 days apart is statistically extraordinary regardless of their sky positions. The question is whether it's coincidence or a shared disruption origin with divergent orbital evolution. The data doesn't prove it either way, but "they came from different directions therefore unrelated" is not the gotcha Claude thinks it is.
The bottom line:
The debunk amounts to: "it's not in the published literature, therefore it's wrong." All new findings are unpublished before they're published. The AMS data is public CSV. The GMN data is freely downloadable from the IAU MDC at ceres.ta3.sk. The analysis is a straightforward autocorrelation. Anyone can reproduce it. If the 27-day signal in 735,000 independent orbits is "sophisticated-sounding confabulation," I'd like to see what disproves it — because "Claude said so" isn't a methodology.
(Score: 2) by driverless on Wednesday April 08, @07:43AM (2 children)
Typical, they can't even hit Narn. Did they hire stormtroopers to run their mass drivers?
(Score: 2) by PiMuNu on Wednesday April 08, @08:11AM (1 child)
Actually they are coming from Mongo.
(Score: 3, Touché) by driverless on Wednesday April 08, @08:17AM
Quite possibly, I hear they got pretty pissed about all the web-scale jokes.
(Score: 1) by RootTwo on Wednesday April 08, @03:03PM
They're getting revenge for NASA's DART mission [wikipedia.org] that impacted their moon Dimorphos.
(Score: 3, Interesting) by VLM on Wednesday April 08, @05:39PM (1 child)
Probably a UFO, more in the terms of a military experiment or drone.
The problem with extraterrestrial origin is "your average meteor" is about 60 miles up when its all glowie.
So if you live in an area with thunderstorms (I do) you grow up counting the seconds between "the flash" and "the thunder" and its about 5 seconds per mile. You can get a pretty good idea of if the storm is headed toward or away from you by looking at the lightning direction and local winds and timing for distance. This is a pretty stereotypical late grade school or very early middle school "get out the protractor and ruler and plot the storm as homework"
So when you get a report of "it flew overhead and went boom" thats gotta be close, low altitude like 5000 feet at most, probably a F-22 or similar shooting down a drone, or F-22 practicing shooting down a drone, or a drone practicing shooting down a F-22 LOL, or similar foolishness. At 60 miles straight up the report should be "it flew overhead and six minutes later there was a boom"
(Score: 2) by VLM on Wednesday April 08, @05:53PM
Oooh more precisely could this have been a task required (or optional) for a late 80s / early 90s Boy Scouts meteorology badge? I honestly don't remember if I got that one. I believe its merely called "weather badge" now.
I looked at scouting.org right now and the modern dumbed down "create a weather log" might have been a little more hard core WRT storm tracking back in the day. Maybe. The protractor, ruler, and graph paper thing might have been my "track a storm for the badge" back in the day.
I vaguely recall based upon being a kid myself and my kids more recent experience that scouting used to be a bit more chill about ".... or agree to do something of similar or equal effort approved by your merit badge counselor." Possibly tracking some rando thunderstorm in the late 80s was my modified version of "make a written weather log".
Alternatively, if you don't program kids every moment and force them into predetermined paths, sure, they'll usually just F around, but once in a while they'll create their own play time that can be pretty cool.
For decades every time I go to scouting.org I think about how adults need something like this, despite, or maybe because of the internet. "A organized path to FAFO". I guess for "computer peeps" there's roadmap.sh or maybe exercism. But you'd think there would be interest in "a small book to teach adults how to ... something"