Antonio Vernucci wrote:
I am simulating on EZNEC an antenna yielding a gain of about 13.5 dB (at
zero deg. elevation) when in free space.

The same antenna, placed on an ideal ground, shows a gain of about 18.8
dB (at 10 deg. elevation), thanks to the so-called "ground gain".

If the antenna is placed on a non-ideal ground (according to EZNEC
modelling), its gain becomes 18.4 dB.

I am surprised of the small gain difference between the ideal ground
case and the non-ideal ground case (just 0.4 dB). As a matter of fact,
while the ideal ground may be regarded as a perfectly conductive plane,
the non-ideal ground is something having a significant ohmic resistance.
I would had thought that most of the RF power incident on a plane having
a significant resistance would be absorbed (and hence not reflected).

Any comment please?

You must be modeling a horizontally polarized antenna. The attenuation
caused by ground reflection is quite different for horizontally and
vertically polarized fields. When the field is horizontally polarized,
nearly any reasonable ground gives a near-perfect reflection at low
elevation angles. Only at high elevation angles will you see much
difference caused by ground quality.

Unfortunately for people using vertical antennas, the story is quite
different when the field is vertically polarized.

There's a good discussion of the effect of imperfect ground and
derivations of the ground reflection coefficients in any edition of
Kraus' _Antennas_.

Roy Lewallen, W7EL