In some sense, it's like modeling ionospheric "reflection" as a single height,
when in reality it's a diffuse refractive effect spanning many 10s-100s of km.
That said, I think one can come up with a frequency specific "single useful
number" given an assumed soil profile. You'd get an "effective depth" and a
rolled up reflection coefficient (which would be angle dependent).
I suspect, also, that for H-pol, it's WAY simpler than for V-pol, just because
the "first surface" is so reflective at low angles.
On Wed, 15 Apr 2026 20:50:30 -0700, David Gilbert via TowerTalk
<towertalk@contesting.com> wrote:
Hi, Brian.
Yeah, I almost always use the phrase "effective ground plane" when I
talk about this stuff but I didn't there. Also, I know that it's
actually a spatially distributed effect, but that gets even clumsier to
say each time ;)
I know that you've done a ton of really good work on ground conductivity
and ground effects and as you say I am no doubt seeing some very complex
effects given the ground conditions I have here ... very dry soil on
top, probably more moist soil down deep, and composition that ranges
from sand and fine particles to rocks the size of a bus at varying
depths. I have a drone that I have once or twice used to plot the
elevation pattern of an antenna (it works surprisingly well), and I may
some day model a couple of antennas and compare their actual measured
elevation plots (using the drone) to see what I really have and how much
that varies across my lot.
Lots of cool stuff to do ...
73,
Dave AB7E
On 4/15/2026 6:16 PM, Brian Beezley wrote:
> AB7E said:
>
> "Keep in mind that the RF ground plane is almost certainly not at the
> surface of the physical ground."
>
> Dave, the notion of a ground plane makes sense for perfectly
> conducting ground. And it's not far off for seawater. But for other
> types of ground, it can be misleading.
>
> Electrical ground is right where you see it. But it's only part of the
> story. For generic desert soil like yours, skin depth is 157 feet at
> 14.2 MHz and 533 feet at 1.8 MHz. That's where antenna-induced ground
> current has decayed to 37% of its surface value. It's still 13.5% at
> two skin depths. A lot can change over such distances, especially
> moisture content, which greatly affects ground permittivity and
> conductivity.
>
> For any type of soil, the Fresnel reflection coefficient is nonzero at
> the air/ground interface. Some signal reflection occurs there. For
> low-loss soil like yours, significant reflection also may occur at
> deeper soil layers. An upward-going signal reflected from a deep
> stratified layer may get re-reflected at the surface. It can bounce
> back and forth within the layer, dissipating power as it travels. This
> effect can occur at multiple layers. Each upward-going signal that
> reaches the surface transmits some power into the air. The result can
> be quite complex. A mathematical treatment is here:
>
> https://msp.org/memocs/2016/4-2/memocs-v4-n2-p03-p.pdf
>
> Some of your inconsistent phase results may stem from subsurface
> reflections with long path lengths. Unwrapping phase may help.
>
> A fascinating effect is the resonance that can occur for low-loss soil
> when the distance between subsoil layers is near a multiple of a
> half-wavelength. I calculated that a sandy aquifer 66 feet below a
> desert surface (a half-wavelength in ground) can increase effective
> surface conductivity by a factor of 30 at 3.7 MHz. I suspect such
> effects seldom occurs in practice because everything has to be just
> right. You can download a calculator to explore resonance and other
> stratified ground effects for two soil layers here:
>
> https://k6sti.neocities.org/sg
>
> Subsurface soil can affect both antenna impedance and far-field
> patterns. NEC models uniform soil only. Its results may not be
> realistic at your QTH because of the large exposure to subsurface
> effects for desert soil.
>
> Brian
>
> _______________________________________________
>
>
>
> _______________________________________________
> TowerTalk mailing list
> TowerTalk@contesting.com
> http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|