The pinnacle of Aguille de la Tsa, casts its morning shadow across a valley. Imaged in the Swiss Valais by James Osborn August '07 from the pinnacle top (see lower insert view) . Large image ©James Osborn, insert summitpost.org.
The shadows have several intriguing features. Why are they several shades of blue? Why is the pinnacle shadow knife-like and inky dark? Why do the shadow edges converge towards the tip of the pinnacle shadow?
We are looking at two superimposed shadow types. The first is the 2D ridge shadow cast onto the opposite valley wall (mentally flip the small insert image left to right to match the shadow features). The second series of shadows are solid, 3D, cast into the intervening air itself. It is these that create the interesting features.
The shadowed air is a dark void. Its top has parallel corrugations corresponding to the ridge profile. Make some parallel folds in a sheet of paper to model it. The camera sees the corrugations from the sharp fold coming from the pinnacle top - the red spot in the diagram.
Shadow blues - Look from the red spot along the shadow of the pinnacle. All the air in the path is in complete shadow and is dark. Look across to the parallel shadow of the adjacent peak. Your line of sight crosses intervening sunlit air. Scattered light from the sunlit air lightens the appearance of the shadow. The next peak's shadow is lightened even more by two bands of intervening sunlit air.
Sharp pinnacle shadow - the dark shadow is in the air. it is a 3D air shadow, not the 2D one on the valley wall. From the red spot the pinnacle's 3D shadow void appears as a sharp wedge shape.
Converging shadow edges - The uppermost ridges of the shadows are all parallel. But from the red spot they appear, by perspective, to converge on the antisolar point directly opposite the sun and marked by the shadow of the pinnacle peak on the opposite valley wall (blue spot). To see the effect 'live' try looking along a parallel folded sheet of paper with your eye at the red spot position.