What causes circular holes in medium-level clouds?

Gordon Tripp asks: "What causes the very neat, circular holes in medium-level clouds showing clear blue sky above? They are periodically seen over the North Downs in Surrey and can last for a considerable time."

My first reaction is that these are the result of seeding of the supercooled water droplets in the medium-level cloud by ice crystals from cirrus clouds above them. You can do this seeding yourself in a chest deep-freeze. You'll need some dry ice (solid CO2). If there's none handy in the lab or school, try this directory.

To see the effect best, set up a projector pointing into a mirror above the open freezer so that a bright beam shines down into the chest. Breathe into the chest. A cloud forms. It continues as liquid droplets even though the temperature in the freezer will be something like -18°C (zero Fahrenheit). Such 'supercooled' water is frequent in the atmosphere. Now drop a small pellet of dry ice into your 'cloud'. It is so much colder than the water droplets that droplets close to its track freeze instantaneously, their outer shell splitting and exploding as the liquid water expands on freezing. In turn, these ice splinters cause their neighbouring drops to freeze and the effect spreads: the whole cloud rapidly turns to one of crystals glittering in the bright beam of light.

The growth of ice crystals at the expense of liquid water drops is the basis of much of our rain in mid-latitudes. The physics is described here. It has been the basis of many attempts to modify the weather. Certainly, supercooled stratocumulus cloud has been seeded from aeroplanes dropping dry ice pellets into it. Dropped along an L-shaped flight path, for example, they cause an L-shaped hole to form in about 20 minutes. The hole stays open for about another 20 minutes and then takes a further 20 minutes to fill up with cloud again.

If the seeding of your medium-level Altoculumus by fallstreaks from Cirrus were continuous, the holes might last much longer. Usually, however, winds at the two cloud levels would be different and so the hole in the cloud would move. If the wind at Cirris levels were calm, though, the hole might stay in the same place relative to the ground.

An entirely different cause might be the North Downs themselves. Holes sometimes occur in stratocumulus cloud when low-level winds sink down the lee side of hills. The sinking air warms by compression as it sinks and causes the cloud to evaporate. See here, for example. However, circular holes in medium-level altocumulus from relatively small hills, such as the North Downs, seem less likely. I favour the seeding idea.

Many of the ideas behind both processes are covered in this American Powerpoint presentation. Circular fallstreak holes, in particular, are discussed here.