[TowerTalk] Snow and rain attenuation

[Jim Lux]

At 11:37 AM 10/31/2005, Robert Chudek wrote:
>Hi Bill,
>
>The rain and snow attenuation seems to always raise this type of question. 
>At the opposite end of the spectrum (440 MHz) where I ran ATV, snow and 
>rain did not have a perceptible influence on the signals either. What did 
>influence the path was foliage. In the winter we could shoot signals 
>through the trees, but in the summer there was severe attenuation. I 
>suspected the issue was the water or sap rising during the growing season.
>
>But that aside, is there a frequency where rain and snow have a 
>significant attenuation due to its presence in the path?

In general, liquid water attenuates more strongly than frozen water, so 
rain is much wose than snow or ice. (but not all snow/ice is dry!).  The 
usual number I've seen where folks start to worry about rain fades is 
around 7GHz, but that might be an artifact of frequency allocations for 
applications where people actually care.  Radar designers have to deal with 
this, too.

In general, attenuation goes up as frequency goes up in a fairly smooth 
curve until you get to around 150 GHz.  Rain attenuation is usually modeled 
as a power law:
  gamma = K*RainRate^a
where K and a are a function of drop size (as a function of wavelength) and 
temperature.  Some specific numbers:

At 10 GHz, 5mm/hr, 0.08 dB/km
At 20 GHz, 5mm/hr, 0.5 dB/km
at 30 GHz, 5mm/hr, 1 dB/km

Another significant effect is that rain (and ice crystals) depolarize the 
signal (in fact, pointing straight up into falling rain is one way to 
generate a evenly depolarized signal scattered back down for testing).

Thre are a couple frequencies where there is very strong absorption due to 
molecular resonance effects (22 GHz for H2O, 60 GHz for O2), but those 
aren't really rain related.

Jim, W6RMK



Jim, W6RMK