Moon illusion

People commonly report that the Moon looks unusually large when it is near the horizon and smaller when it is high overhead. The apparent enlargement is a perceptual effect rather than a substantial change in the Moon's angular size. A simple comparison can be made by holding a fingertip at arm's length against the Moon at different elevations; photographs taken with unchanged camera settings likewise show very similar widths. The apparent size can be especially striking when trees, buildings or mountains are in view.

The Moon is not physically closer to an observer at the horizon; because of Earth's geometry, it is slightly farther away there than at the zenith. Atmospheric refraction also does not magnify it. Refraction near the horizon can make the lunar disk appear somewhat vertically flattened, while atmospheric scattering may make it look more yellow, orange or red. These atmospheric effects are distinct from the impression of a larger disk.

Many accounts focus on how visual perception combines the image on the retina with estimates of distance and surrounding context. In apparent-distance or size-constancy accounts, the brain may interpret a horizon Moon as farther away, and a familiar size viewed as farther away can be experienced as larger. Foreground objects and perspective cues may contribute in some settings; the Ponzo illusion is often used as an analogy. Experimental work has tested these ideas, including studies in which artificial Moon-like targets were viewed against different visual backgrounds. Findings support particular perceptual mechanisms in some circumstances but do not settle the full explanation.

The Moon illusion has been discussed for centuries, but its precise causes remain debated. NASA's overview notes that there is no rock-solid explanation and that more than one factor may be involved. The illusion is also reported in settings without ordinary foreground landmarks, indicating that a single simple account based only on horizon scenery is incomplete. It is therefore more accurate to describe proposed explanations as hypotheses than to present one as established fact.