[TowerTalk] antennas in the forest

[Jim Lux]

Reading on, especially the papers by Tamir (1967 in particular) and 
Cavalcante..

If your antenna is buried in the forest (i.e. it's below the tops of the 
trees) then the ground parameters of interest are those of the forest, and 
the actual earth parameters are less important.  epsilon is typically 1.05 
- 1.15, so there is a refractive effect in the propagation. Sigma 
(conductivity) is typically 0.05-0.15 mSiemens/meter, so it's not 
particularly lossy, compared to earth.  You can compute the "refractive 
index" of the forest for ray/optical kinds of computations. Tamir gives the 
following:

abs(n^2-1) = abs( epsilon - 1 -j*60*sigma*lambda) = approx 60*sigma*lambda

Given a sigma of 0.1E-3 and 20m, I get abs(n^2-1) = .12, or n = 1.06.

(Using Snell's law, sin(critical angle) = 1/n, tells you when total 
reflection will occur. For that n=1.06, it means that if the angle 70 
degrees from vertical (i.e. 20 degrees elevation or closer to horizon), 
then you'll meet the condition for total reflection from the interface... 
this is pretty significant, guys.. the wave at low angles is going to 
propagate within the forest, like in a waveguide...)

Note that if your antenna is anywhere close to the forest/air interface, 
this also applies.  Just getting your antenna 10 feet above the tree tops 
doesn't remove the effect of the forest (and, actually, might make it 
worse..).  It's much like antennas over ground, where the properties of the 
ground affects the pattern.

A lateral wave develops, propagating along the interface between the tops 
of the trees and the air above (similar to a "ground wave" familiar to AM 
MW listeners), so the foliage attentuation for short paths (<100km) is 
independent of distance (in other words, the path is not through the 
trees), and you just add in a "foliage factor" for loss. There are paths 
contributing to this lateral wave that reflect from the forest/air boundary 
as well as the forest/ground boundary (obviously, the hop distance is quite 
short, and this occurs even if the wavelength of the signal is greater than 
the height of the forest.. the math just gets more complex, but, in this 
case (long wavelength) the effect is much less.)

For skywave paths with low elevation angles, the lateral wave contributes 
significantly to the total signal transmitted or received.

Forested areas have much less effect from the roughness of terrain.  The EM 
properties of the forest dominate. (This is relevant for users of programs 
like HFTA, which assume  a uniform earth EM properties and do ray tracing)

There is an interesting effect where the optimum orientation of the antenna 
is somewhere between vertical and horizontal, but knowing what it is 
depends a LOT on the exact EM properties of the forest.  The effect is on 
the order of 0.5 to 3 dB.

Forest depolarizes the transmitted wave.

Here's an observation which DX'ers might be interested in:

Forest might reduce fading depth on skywave paths, because it makes the 
antenna effectively "larger", so it averages more paths. (of course, you'll 
also have some additional loss from the forest, so the peaks will be lower, 
too)


Anyway, it looks like forests and trees DO have a significant effect on HF 
propagation.

None of this, by the way, addresses the effects of trees in the "near 
field" (say, within 2*L^2/lambda of the antenna.. ), but those effects are 
probably more in the nature of added loss.

Jim, W6RMK