The dark shade of the moon

Abstract

reason moon luminance increases rapidly with altitude, from 2 cd/m 2 up to over 4000. 3 Accordingly, the moon can appear off-white, white, fluorescent, or luminous – variations that are not predicted by Retinex theory. Both the colour and the mutable appearance of the moon are explained well by models based on luminance anchoring within a visual scene, such as the double-anchoring model of lightness. 4,5 In this model, the shade of grey (technically, the ‘lightness’) of any given region is computed by taking a weighted average of the ratios of the region’s luminance to two ‘anchors’: the surround luminance and the highest luminance in the scene. Such anchors are both given a default value of white. As the brightest object in the night sky, at the highest-luminance anchoring stage the moon is always perceptually white, whatever its actual luminance. However, at the surround anchoring stage, it is lighter than white (and thereby perceptually glowing), and the extent of this component depends on the moon’s luminance ratio to the surrounding sky. From the standpoint of anchoring models of lightness, the shade of the moon is a spectacular, but otherwise predictable, instance of the general principle that the achromatic colour of objects results from luminance anchoring within the visual scene.