Pressure depresses the melting point of ice, so (as pointed out in the article) it is not a candidate explanation for ice remaining underneath the coin. Furthermore, if the pressure is insufficient to depress the melting point below the ambient temperature, it has no effect.
The claim that the highest pressure is just under the center of gravity does not hold in general, either. More generally, the pressure distribution depends on the distribution of the contact points between the body and its support (consider a four-legged table, for example.)
The article does not say that the umbrella shade hypothesis explains the phenomenon, though it is a factor, as it affects heat flow. The article suggests (very plausibly) that it is due to sublimation. As this only occurs where the ice is in contact with the air[1] (and that air is free to move and carry away the sublimated vapor), the erosion proceeds inwardly from the boundary of contact.
For single pedestals, the boundary does have to shrink in such a way that it always contains the point below the body's center of gravity, so I would like to see the experiment conducted with irregularly-shaped bodies (if I had more free room in my freezer, I might try it myself.) Hypothesis: for bodies of uniform thickness and density, this is often the case.
[1] As demonstrated by the apparatus used in the experiment (where the sublimation occurred into a vacuum), what really matters is exposure of the ice to the void between the molecules of the air.
The claim that the highest pressure is just under the center of gravity does not hold in general, either. More generally, the pressure distribution depends on the distribution of the contact points between the body and its support (consider a four-legged table, for example.)
The article does not say that the umbrella shade hypothesis explains the phenomenon, though it is a factor, as it affects heat flow. The article suggests (very plausibly) that it is due to sublimation. As this only occurs where the ice is in contact with the air[1] (and that air is free to move and carry away the sublimated vapor), the erosion proceeds inwardly from the boundary of contact.
For single pedestals, the boundary does have to shrink in such a way that it always contains the point below the body's center of gravity, so I would like to see the experiment conducted with irregularly-shaped bodies (if I had more free room in my freezer, I might try it myself.) Hypothesis: for bodies of uniform thickness and density, this is often the case.
[1] As demonstrated by the apparatus used in the experiment (where the sublimation occurred into a vacuum), what really matters is exposure of the ice to the void between the molecules of the air.