How to evaluate ground reflections in Pattern testing?

[phased]

I am worried that I may be suffering from significant ground-reflections in my Antenna Patterns. -I am Pattern testing horn antennas at around 1GHz. The source horn has a fairly broad beamwidth. (so too do the Horn_Under_Test). I am Pattern Testing in my back garden in Rural Cumberland (very remote). I have limited resources and just have to mount the Source and Test horns 2 metres above the grassy ground.
Would you agree that in order to evaluate the effect of reflections, the following would need to be done:
(1) Peak the horns and then gradually move the horn under test directly backward from the Source Horn.
(2) Move the horn_under_test backwards over sufficient distance such that it definetely would have encountered a ground_reflected ray which was an integer number of half-wavelengths different in length than the direct ray.
(3) Observe for deep "dips" in received power as the horn_under_test is moved backward -If deep "dips" in received power are seen, then conclude that reflections are significant.

May i enquire as to whether this is a good technique for cheaply investigating the effect of reflections on my set-up?

 

[VEBill]

There are plenty of variables that you can change while holding the true (pattern) variables constant.

Some examples:
-Lower the ground ;-) , or raise the entire test set-up
-Tilt the entire test set-up (while holding pattern angles constant
-Position a flat reflector midway and angled to send the bulk of the indirect wave into outer space
-Slightly vary the frequency such that the primary effect is that the summation goes in and out of phase while the change is sufficiently slight that the antennas' patterns wouldn't change very much

You can pretty much see the reflections in the pattern in the frequency domain. The amplitude is obvious and the frequency domain periodicity tells you the relative phase changes (two paths, different lengths) with wavelength.

If I were doing this in my back yard (garden), I'd try to aim the source straight up (as much as possible) and suspend the UUT.

Disclaimer - I haven't done this in my bard yard, but I've been involved with similar tests at work and I've seen reflections in the patterns (very obvious).

 

[BrianR]

I once had to measure the pattern of a 1GHz antenna mounted on the planar radar plate on the nose of an aircraft and encountered exactly the conditions you are facing. I mounted the horn on a cherry picker and varied its height above ground rather and moving it in and out and encountered +6db/-20db fluctuations due to ground reflections. In the end we found a stable area by experiment which needed the aircraft antenna facing as far skyward as it would go (as suggested by VE1BLL). Even facing vertically upwards you will get ground reflections from antenna sidelobes.

You can build sawtooth fences and put them on the ground in between the two antennas but experiment will certainly be needed as ground reflections will be significant.