A 0.8-Meter Asteroid Got 7 Hours of Warning. Scaling the Physics Gives Chelyabinsk 8 Days.

On Sunday, September 6, an automated telescope on Mount Lemmon in Arizona picked up a faint speck moving against the background stars. Seven hours and twenty-two minutes later, that speck burned up over the Indian Ocean at 16:07 UTC. It was eighty centimeters across, about the size of a beach ball, and completely harmless.

Why does it matter? Because it was only the 13th time in human history that anyone saw an asteroid coming before it hit the atmosphere. Across eighteen years of trying, the total is thirteen. And because 2026 RW1 happened to arrive on a clean, detectable trajectory, we can run a calculation nobody has published: scale the detection physics to a bigger rock on the same path, and a Chelyabinsk-class impactor, the kind that injured 1,600 people in 2013, would have bought roughly eight days of warning instead of none at all.

The 13 ever caught

ESA's Planetary Defence Office keeps the official ledger, and it is a short document. 2008 TC3 came first, spotted by this same Mount Lemmon Survey 20 hours before it scattered meteorites across the Nubian Desert in Sudan. Then the trickle: 2014 AA, 2018 LA, 2019 MO. Then the pace quickens, because nine of the thirteen were caught between 2022 and this week.

In January 2024, Kitt Peak found 2024 BX1 less than three hours before it lit up the sky over Berlin. All of them, with discoverers and warning times measured in fractions of a day, sit in Table 1 of a 2025 arXiv paper forecasting how many imminent impactors the Vera Rubin Observatory's LSST should add to the count, and the table reads like a sport that keeps getting faster: four detections in eleven years (2008 to 2019), then nine in under five.

That is roughly a fivefold acceleration, driven by faster surveys: ATLAS in Hawaii, the Konkoly Observatory in Hungary, and the tireless Mount Lemmon cameras. Raw counts flatter the picture, though. What matters is the fraction of incoming rocks we catch, and for that you need the denominator.

The denominator is enormous

The standard flux estimate comes from Brown et al. (2002), who used eight years of U.S. satellite optical-flash data to fit a power law: N(>D) = 37 per year times (D/1 m) to the power of negative 2.7. Run the numbers and the sky turns out to be raining gravel:

Impactor size	Impacts per year	2026 RW1's league
> 0.7 m	~97	All 13 detections
> 0.8 m	~68	RW1-sized, ~0.8 m
> 1 m	~37	Car-sized and up
> 20 m	~0.011 (1 per ~88 yr)	Chelyabinsk-class
> 140 m	~0.00006 (1 per ~17,000 yr)	Congress's "city killer" line

Thirteen catches in eighteen years, against roughly 97 rocks a year at the 0.7-meter threshold, works out to about 0.7 percent. Even 2024, the best year on record with four catches, managed roughly 4 percent. We can celebrate the catch without trusting the system.

Scaling the warning

Here is the calculation that makes this rock more than a curiosity. Detection range scales with an object's diameter. Reflected sunlight grows as the square of the diameter while the inverse-square law dilutes it with distance, so for a fixed limiting magnitude the telescope sees a bigger rock proportionally farther out, and farther out at the same closing speed means proportionally more warning.

Start from RW1: 0.8 meters, 7.37 hours of warning. A 20-meter Chelyabinsk-class rock on the identical path would be visible 25 times farther away, for 25 times as long: 184 hours, about 7.7 days. A 140-meter object, the size Congress ordered NASA to catalog 90 percent of, gets 175 times the warning: roughly 54 days. Eight days is enough to close schools and board up windows. Fifty-four days is enough to do something about it.

These are best-case numbers along a clean geometry, and the caveats bite hard. Surveys revisit each patch of sky only every few days, so a rock can brighten past the threshold between visits. Albedo varies, so a coal-dark rock stays dimmer than a bright one of the same size. And none of this scaling helps against an object coming from the direction of the Sun, where ground telescopes cannot look at all. The physics is generous; the geometry is not.

Chelyabinsk is the counterexample

The strongest case against this kind of optimism is Chelyabinsk itself. In February 2013 a rock about 20 meters across exploded at roughly 25 kilometers altitude with an energy of 20 to 30 times the Hiroshima bomb, and more than 1,600 people were injured, mostly by flying glass, with roughly $30 million in property damage done. Approaching from the daytime sky, it gave no ground-based optical survey a chance. Warning: zero.

That exact blind spot is what NASA's NEO Surveyor is being built to close. It is a 50-centimeter infrared telescope stationed at the Sun-Earth L1 point, sensing heat instead of reflected light, which sees dark rocks that visible light misses and stares into the daytime sky that ground telescopes cannot scan. It launches no earlier than September 2027, after its 2026 date slipped and its budget doubled to $1.2 billion, with a mandate for a five-year baseline survey to find at least two-thirds of near-Earth objects larger than 140 meters, working toward Congress's 90 percent goal.

Limitations

Now the honest accounting. Everything above rests on the Brown et al. 2002 flux law, fitted to satellite flash data from 1994 to 2002 and extrapolated by a power law well past its calibration range; modern re-estimates move the numbers by up to a factor of two, which shifts the 0.7 percent figure but not its order of magnitude. The warning-time scaling assumes identical trajectory, speed, albedo, and phase geometry, plus a fixed survey limiting magnitude, so treat 7.7 days and 54 days as physics upper bounds, not system guarantees. Real warning time is set by survey cadence and approach geometry. ESA's 0.6 to 1.3 meter range for RW1 is inferred from absolute magnitude under an assumed albedo, not measured directly. And 2026 JN4, the May 2026 impactor that brought the ledger to twelve, has thin public reporting, so this analysis leans on the twelve with full published records.

What to watch

The metric to watch is the pre-detection fraction: under 1 percent all-time, about 4 percent in the best year. Watch whether the next predicted impactor is caught by a ground survey or from space. Every ground catch is a win for cadence and coverage. One catch from a space-based infrared survey, and the sunward blind spot is finally closing. And watch NEO Surveyor's launch date, because the Chelyabinsk gap is one infrared telescope at L1 away from being addressed.

The Bottom Line

We caught a beach-ball-sized rock seven hours early, and the physics of that catch implies roughly a week of warning for a Chelyabinsk and nearly two months for a city killer, provided they arrive on a clean path. In reality we still catch fewer than 1 in 100 of the rocks that actually arrive. The detection rate is accelerating fivefold per decade and the ledger just hit thirteen, but the real insurance policy is a single infrared telescope parked a million miles sunward, currently waiting on a 2027 launch. Until it flies, the most dangerous approach vector is the one we literally cannot see.

Sources

ESA Planetary Defence Office: Past Impactors (official ledger of pre-impact detections with diameters, velocities, and energies)
"Predictions of Imminent Earth Impactors Discovered by LSST" (2025) (Table 1: all 11 pre-2025 detections with discoverers, warning times, diameters)
Tonry et al. 2011, "An Early Warning System for Asteroid Impact" (ATLAS design paper) (quotes Brown et al. 2002 flux law N(>D) = 37 yr⁻¹ (D/1m)^−2.7)
NASA JPL: Near-Earth Object Surveyor (launch no earlier than September 2027, 140 m / 90% mandate)
The Watchers: Asteroid 2026 RW1 impacts Earth over Indian Ocean (7h22m discovery-to-impact, 0.6 to 1.3 m estimate, Meerkat impact point)
BBC Sky at Night Magazine (discovery narrative, Aten classification)
National Academies: Finding Hazardous Asteroids (Chelyabinsk damage figures, ~10 million NEOs >20 m)