"Roy Lewallen" wrote
Even if you could make a zero-loss connection to ground (and a large
radial field comes close enough for nearly all practical purposes), that
dirt still doesn't provide the "mathematical image" of a perfect ground
plane.
___________

Expanding on this for S. Wolfe, even if a point connection having zero
resistance to earth potential existed at/near the base of a vertical
monopole, a monopole using that ground reference would be a poor radiator.

This is because the r-f ground currents that need to flow back into the
antenna system first would need to travel through the lossy earth from
distances up to 1/2 wavelength from the monopole, to reach that perfect
ground connection.

The function of the buried radials is to provide a low-resistance path for
those ground currents, which means that they have to be collected as closely
as possible to their sources in the earth (eg, within a disc having a radius
of 1/2-wavelength around the monopole).

A benchmark field study in 1937 by Brown, Lewis and Epstein of RCA showed
that about 120 buried radials each at least 1/4-wave long enable a
groundwave field to be radiated by a monopole that is within a few percent
of that over a perfect ground plane. Ground conductivity at their test site
was no better than 4 mS/m. Their test frequency was 3 MHz.

RF