Impact risk

No known asteroid larger than 140 m across has a significant chance of hitting Earth in the next 100 years.^[1] Newly discovered asteroids do sometimes appear with a small chance of impact, and the most likely outcome by far is that the chance drops away once more observations pin down the orbit.^[2] This guide explains how those chances are worked out, how they are rated, and why they change.

What Sentry does

Sentry is the impact-monitoring system run by NASA's Center for Near-Earth Object Studies (CNEOS) at the Jet Propulsion Laboratory. It is highly automated and scans the catalog of known asteroids for possible Earth impacts over the next 100 years.^[3] Every day it receives new observations and orbits from the Minor Planet Center, and it computes a fresh orbit for an asteroid whenever new telescope or radar measurements arrive.^[2]

Sentry then follows each asteroid's motion at least 100 years ahead, allowing for the gravity of the Sun, the planets, the Moon and 16 of the largest asteroids.^[2] An orbit is never known exactly, so Sentry does not follow one path. It follows many virtual asteroids, each a slightly different orbit that still fits the observations, and looks for any that collide with Earth.^[2] When it finds some, the asteroid goes on the Sentry risk table, with each possible impact date rated on the Torino and Palermo scales.^[2]

Impact probability

An impact probability is the share of those possible orbits that end in a collision. If 100,000 virtual asteroids are followed and 2 of them hit Earth in a given year, the impact probability for that year is about 1 in 50,000.^[2] The number describes what is known so far, not a fixed property of the asteroid.

Why the odds usually fall to zero

A newly found asteroid has been seen along only a short stretch of its path, so its future position is uncertain over a wide region of space. If Earth lies inside that region on some future date, the asteroid has a small impact probability for that date.^[2] Each new night of observations shrinks the region, and for the probability to reach zero, Earth has to fall outside it.^[4]

While the region is still shrinking around Earth, the probability can rise before it falls; ESA notes that this pattern is common, with the probability often dropping quickly to zero after further observations.^[5] Asteroid 2024 YR4 is a recent example: discovered on 27 December 2024 by the NASA-funded ATLAS telescope in Chile, it was estimated at 40 to 90 m across.^[6] Its chance of hitting Earth on 22 December 2032 passed 1% on 27 January 2025.^[6] On 18 February it reached 3.1%, the highest NASA had ever recorded for an object of that size or larger, and the next day new data lowered it to 1.5%.^[4] By 24 February it stood at 0.004%.^[7]

Apophis went through the same sequence two decades earlier. Soon after its discovery in 2004, it was briefly estimated to have a 2.7% chance of hitting Earth in 2029.^[8] Further observations ruled out 2029, then a possible impact in 2036, and radar measurements in March 2021 ruled out the last remaining date, in 2068.^[9] NASA now finds no risk of an Apophis impact for at least 100 years.^[9]

What "removed from the list" means

When new observations leave no possible impacts, Sentry removes the asteroid from its risk table. Several newly found asteroids each month are listed and then removed shortly afterwards; CNEOS describes this as a normal, expected process.^[2] Removal does not mean the first assessment was a mistake. The risk was real, given what was known, until additional observations showed that it was not.^[2]

A few objects stay on the table for years. These were observed only briefly before they were lost from view, and they may be removed only if they happen to be found again.^[2]

The Torino scale

The Torino scale was adopted by the International Astronomical Union in 1999 to communicate impact predictions to the public.^[10] It rates a possible impact from 0 to 10, taking in both how likely it is and how much damage it could do, but not how many years away it is.^[10]

• Level 0, white: no hazard. The chance of collision is zero or so low as to be effectively zero.^[10]
• Level 1, green: normal. A routine discovery of a pass near Earth, with no cause for public concern; new observations will very likely lower it to level 0.^[10]
• Levels 2 to 4, yellow: meriting attention by astronomers. Level 2 is a somewhat close but not highly unusual pass.^[10] Level 3 means a 1% or greater chance of a collision capable of localized destruction, and level 4 the same for regional devastation; both merit public attention if the encounter is less than a decade away, and both will most likely be lowered to level 0.^[10]
• Levels 5 to 7, orange: threatening. A serious but still uncertain threat of regional devastation (level 5) or of a global catastrophe (levels 6 and 7), calling for critical attention from astronomers and, depending on how soon the encounter is, government or international contingency planning.^[10]
• Levels 8 to 10, red: certain collisions. Level 8 is localized destruction, an event that happens on average between once per 50 years and once per several thousand years; level 9 is unprecedented regional devastation, between once per 10,000 and once per 100,000 years; level 10 is a global climatic catastrophe, once per 100,000 years or less often.^[10]

Apophis reached level 4 in December 2004, when its chance of impact in 2029 rose to about 1.6%.^[11]

The Palermo scale

The Palermo scale is the scale specialists use to compare and prioritize possible impacts. It measures a predicted impact against the background risk: the average chance that some asteroid of the same size or larger hits Earth in the years before that date.^[12] Because it counts those years, it gives a near-term threat more weight than a distant one.

The scale is logarithmic. A value of 0 means the predicted impact is as threatening as the background, -2 means it is only 1% as likely, and +2 means it is 100 times more likely.^[12] Values below -2 reflect events with no likely consequences, values between -2 and 0 merit careful monitoring, and positive values generally merit some concern.^[12] Positive values are very rare: almost every entry on the Sentry table sits below zero, meaning less risk than the background that exists whether or not the asteroid is known.^[2]

Reading a risk listing

An entry on the Sentry table means an orbit is still uncertain, not that an impact is expected. Even the most closely studied cases carry small numbers: Bennu, one of the two most hazardous known asteroids, has a total chance of about 1 in 1,750 of hitting Earth through the year 2300.^[13] For how close asteroids actually come, and what an impact of each size would do, see How close, how big, how often.