Richard Harrison wrote:
. . .
It should work at r-f too except
for skin effect which if I recall causes an increase in resistance
proportional to the square root of the frequency. The skin thickness is
proportional to the reciprocal of the square root of the frequency. . .

That's an approximation which is true only for a good conductor. A poor
conductor like ground acts like a conductor only up to a particular
frequency, above which it acts like a dielectric in which the skin depth
stays constant with frequency. Of course, there isn't an abrupt change,
but the characteristic transitions from one to the other much like the
frequency response of a highpass filter transitions from a fixed rising
characteristic to flatten at the break frequency.
Here are the transition frequencies for some common types of ground,
where the first number in parentheses is the conductivity in S/m, the
second is the relative permittivity (dielectric constant):

Very good (0.0303, 20): 27.2 MHz
Average (0.005, 13): 6.9 MHz
Poor (0.002, 13): 2.8 MHz

For many purposes involving electromagnetics including determining skin
depth, the ground acts like a conductor below the transition frequency
and like a dielectric above it.

For example, here's the skin depth in feet for average soil at various
frequencies:

Freq MHz Skin Depth Ft

0.5 34.2
1 25.1
3.5 15.9
7 13.8
10 13.2
30 12.6
100 12.6

A very good treatment of this can be found in Kraus' _Electromagnetics_.

Roy Lewallen, W7EL