A tiny asteroid can be too faint to study in reflected sunlight yet still interrupt a much brighter background star. A stellar occultation uses that interruption to measure a silhouette. Its advantage is geometric: the asteroid need not be resolved into a disk. Its difficulty is temporal: the signal may last only a fraction of a second.
Two experiments that should not be confused
In a targeted observation, the asteroid is already known. Its orbit and a star position predict where a shadow might cross Earth. Observers distribute stations around that uncertain path and look for a timed event.
A blind survey instead monitors many stars for unknown objects crossing them. Such a survey needs statistical detection criteria, realistic artificial-signal recovery tests and strong rejection of detector and atmospheric artifacts. A candidate dip alone does not supply an orbit. Distant Kuiper belt occultation surveys are examples of this second approach; their targets should not be described as ordinary main-belt asteroids.
How fast is fast enough?
As a geometric teaching example, a 1 km chord moving at 20 km per second lasts 0.05 seconds. A one-second exposure would average that interruption with much more unblocked starlight. A fast camera helps, but shorter exposures also collect fewer photons. Star brightness, read noise, dead time and accurate timestamps determine whether the event is measurable.
The duration times the projected shadow speed gives a chord length when a geometric model is appropriate. It does not automatically give a diameter: the station may cross near the edge. Several stations constrain the outline far better than one.
When the sharp-shadow approximation breaks down
Light diffracts around an occulting body. A useful scale is F = √(λD/2), where λ is wavelength and D is approximately the object’s distance from the observer for a distant background star. At λ = 0.5 micrometers and D = 1 astronomical unit, F is about 190 meters. At 40 astronomical units it is about 1.2 km.
These are illustrative scales, not hard detection limits. When the object is comparable to this scale, the observed signal can contain diffraction structure rather than a simple rectangular disappearance. Finite stellar diameter, wavelength coverage and exposure averaging must be included. A fitted diffraction model can constrain an event; a stopwatch conversion alone cannot.
A 2024 study of the known asteroid 2001 CC21 provides a concrete small-asteroid example. For the distinct distant-population problem, see the analysis of Kuiper belt occultation diffraction and survey sensitivity.
What makes a brief dip credible?
Check the original images for clouds, cosmic rays, tracking shifts and detector anomalies. Compare nearby stars. Preserve the full time series and calibration information. Independent stations with consistent timing and geometry are especially valuable. A non-detection constrains only signals the equipment could actually have recovered.
An occultation measures a projected shape at one rotational orientation. It does not directly establish mass, density or surface composition. Those require other observations or additional dynamical information.
Continue with why a near-edge chord is short and how secondary dips are tested for rings.
Revised October 26, 2026. The duration and Fresnel calculations are illustrative.