AI-generated illustration for Galileo Whispers. Generic asteroid concept; not a named object, close-approach geometry, orbital solution or impact prediction.
An asteroid’s close approach is predicted by fitting an orbit to observations and calculating how both the asteroid and Earth will move. The result includes uncertainty, because the measurements do not identify the asteroid’s position and velocity perfectly. A predicted miss distance is therefore most useful when read with its uncertainty and the information used to determine the orbit.
This is a physical-distance prediction, not merely a forecast that two objects will look close together in the sky. Our explanation of conjunctions versus close approaches separates those two meanings of “near.”
A few sky positions do not make a perfect orbit
An optical observation measures a direction toward an asteroid at a recorded time. The observer’s own position matters because Earth rotates and travels around the Sun. A sequence of measurements constrains the moving object’s trajectory, rather than providing one fixed location on a celestial map.
Orbit determination adjusts a trial trajectory until its predicted observations agree as well as possible with the measured ones. The remaining differences are called residuals. As JPL’s Sentry introduction explains, the best-fitting trajectory is the nominal orbit. Nearby alternatives can still be compatible with the measurement errors.
Those alternatives represent uncertainty about one asteroid. They are not evidence that the object occupies several places or randomly chooses a new path. The underlying motion follows physical forces; our knowledge of its starting state has limits.
Follow Earth as well as the asteroid
A close approach concerns the separation between two moving bodies. Knowing where the asteroid’s path passes through a region of space is insufficient unless we also know when it gets there and where Earth will be then.
Computers propagate the orbit forward using equations of motion, including relevant gravitational perturbations. A close planetary encounter can change the trajectory significantly and make later uncertainty harder to represent with simple approximations.
NASA JPL’s December 2021 account of impact monitoring describes why accurate modeling must sometimes include small nongravitational effects as well. These are modeled forces, not reasons to portray asteroids as unpredictable objects darting through space. The important task is to carry uncertainty through the calculation rather than erase it from the displayed answer.
Read nominal distance and minimum distance differently
The nominal close-approach distance comes from the best-fit orbit. A listed minimum distance instead asks how close a trajectory within a specified uncertainty boundary could come. These are different summaries of the same prediction problem.
The CNEOS explanation of close-approach uncertainties defines its minimum using the separation between a three-sigma target-plane error ellipse and Earth’s surface. That definition belongs to that product. It is not a universal convention for every distance printed in an astronomy article.
The target plane is a geometrical way to examine the incoming encounter. The uncertainty ellipse describes a boundary in that plane under the adopted model. Its edge is not an observed asteroid surface, and the quoted minimum is not a second, equally certain forecast competing with the nominal distance.
A three-sigma boundary is also not an absolute guarantee that every conceivable trajectory has been included. Its interpretation depends on the uncertainty model. A reader should not turn the word “minimum” into a claim of mathematical certainty about the future.
Why the observation arc matters
CNEOS identifies the number, quality, geometry, and time span of observations as important inputs. The time span is called the data arc. Repeated measurements clustered into one short interval can leave motion less well constrained than observations spread across a longer portion of the orbit.
Consider two hypothetical objects reported with the same nominal miss distance. One was tracked for only a short interval; the other has reliable observations covering years. Their nominal distances alone do not tell you that the predictions have equal confidence. The longer arc can strongly constrain which trajectories remain plausible, although measurement quality and geometry still matter.
New optical or radar observations, when available, can improve the orbit determination. Optical observations can extend the arc as well as add measurements to the fit. These additions can move the nominal solution and shrink its uncertainty. A revised number does not automatically mean the asteroid itself unexpectedly changed course.
A close-approach list is not an impact forecast
Impact monitoring asks a further question: do any trajectories compatible with the observations intersect Earth, and how much probability belongs to those possibilities? That requires more than comparing a nominal miss distance with a headline’s preferred definition of “close.”
The 2021 Sentry-II description illustrates this by examining possible trajectories across an uncertainty region. These computational alternatives are often called virtual asteroids. They help evaluate possibilities for one real object, not a swarm of newly discovered companions.
Before interpreting a close-approach entry, check the object identity, approach date and time, units, distance reference, uncertainty, and orbit-solution information. A new observation can change the assessment. A very precise date without that context may look more informative than the underlying data justify.
There is no actual asteroid alert in the hypothetical comparison above. The useful skill is learning which question each number answers: the best-fit passage, the range supported by an uncertainty model, or the probability of an impact. Those conclusions should never be treated as interchangeable.
