Asteroid impact risk: what's actually being tracked
Nothing catalogued currently threatens Earth — and the story of 2024 YR4 shows the system working.
4 min read · Published 25 August 2026
The short answer
Surveys have catalogued tens of thousands of near-Earth asteroids, and no known object poses a significant impact risk to Earth in the foreseeable future. Risk is communicated using the Torino scale, running 0 to 10. The recent case of 2024 YR4 illustrates the process well: its Earth impact probability briefly rose to around 3% before additional observations eliminated it entirely, which is the normal and expected pattern.
How the process works
A newly discovered asteroid has a poorly constrained orbit, because it has only been observed over a short arc. That uncertainty is represented as a region of possible future positions, and if Earth falls anywhere inside that region, a non-zero impact probability is calculated.
As more observations come in, the uncertainty region shrinks. Usually Earth ends up outside it, and the probability drops to zero.
Counterintuitively, probability often rises first. If Earth is inside a shrinking uncertainty region, it occupies a larger fraction of it as the region contracts — so the calculated probability increases until the moment Earth falls outside, when it collapses to nothing.
This is why headlines about rising asteroid probabilities are usually describing a system working correctly rather than a growing danger.
The Torino scale
| Level | Meaning |
|---|---|
| 0 | No hazard — the vast majority of objects |
| 1 | Normal — a routine discovery with a very low probability, usually reassigned to 0 |
| 2–4 | Meriting attention by astronomers; close encounters deserving monitoring |
| 5–7 | Threatening — serious but still uncertain encounters |
| 8–10 | Certain collision, with increasing scale of consequence |
Almost everything ever catalogued sits at 0. Objects occasionally reach 1 and are downgraded within days or weeks as observations accumulate. Nothing has ever been assigned above 4, and that assignment was later reduced to 0.
What is being watched for
- Near-Earth objects — asteroids and comets whose orbits bring them within a defined distance of Earth's. Tens of thousands are catalogued.
- Potentially hazardous asteroids — a subset that come close enough and are large enough to matter. This is a bookkeeping category, not a prediction of impact.
- Size matters enormously. Objects a few metres across enter the atmosphere regularly and burn up. Tens of metres can cause local damage. Above a kilometre, effects become global — and the survey of kilometre-scale objects is now thought to be substantially complete, with none on a collision course.
- The remaining gap is smaller objects, in the tens-to-hundreds-of-metres range, which are far more numerous and much harder to detect. These could destroy a city without threatening civilisation, and they are the focus of current survey work.
Could anything be done?
Yes, and it has been tested. The DART mission launched in November 2021 and deliberately collided with the small asteroid moonlet Dimorphos in September 2022, measurably changing its orbit around its companion Didymos.
That was the first demonstration that a kinetic impactor works — that hitting an asteroid with a spacecraft can change its trajectory by a measurable amount. The change was small, which is the point: deflection requires only a tiny velocity change if applied years or decades in advance.
The critical variable is warning time. A decade's notice makes deflection plausible. A few months' notice makes evacuation the only option. This is why survey work matters far more than any deflection technology — you cannot deflect what you have not found.
Keeping it in proportion
Impacts large enough to cause regional damage occur on timescales of centuries to millennia. Civilisation-threatening impacts occur on timescales of tens of millions of years.
This is one of the few natural hazards that is genuinely well-characterised, actively surveyed, and in principle preventable with existing technology. It also receives attention out of proportion to its probability compared with hazards that are far more likely.
The honest summary: it is a real risk, it is taken seriously, the current catalogue contains nothing threatening, and the main work remaining is finding the smaller objects. Anything you read implying an imminent known threat is not describing the catalogue.
Common questions
- Is any asteroid going to hit Earth?
- No known object poses a significant impact risk in the foreseeable future. Tens of thousands of near-Earth asteroids are catalogued, the survey of kilometre-scale objects is thought to be substantially complete, and none is on a collision course.
- What is the Torino scale?
- A 0 to 10 scale for communicating asteroid impact risk. Almost everything sits at 0, objects occasionally reach 1 before being downgraded, and nothing has ever been assigned above 4 — an assignment later reduced to 0.
- Why do asteroid impact probabilities go up before going down?
- Because a new object's orbit is uncertain, and if Earth lies inside the region of possible positions, it occupies a larger fraction of that region as the region shrinks with more observations. The probability rises until Earth falls outside it, then drops to zero.
- Can we deflect an asteroid?
- The DART mission demonstrated it in 2022 by colliding with the asteroid moonlet Dimorphos and measurably changing its orbit. Deflection needs only a small velocity change if applied years in advance, which makes warning time — and therefore survey work — the critical factor.