Towertalk
[Top] [All Lists]

[Towertalk] Models and real world

To: <towertalk@contesting.com>
Subject: [Towertalk] Models and real world
From: W8JI@contesting.com (Tom Rauch)
Date: Wed, 10 Jul 2002 08:40:46 -0400
> W8JI wrote:
> >In my experience models tend to fall in line more on 40 meters and up
> >with real-world results, while on 160 results vary the most.
> 
> N4ZR wrote:
> >I believe that most models - certainly prop models like VOACAP --
> >specify 
> that they are not accurate below 2 MHz.  I suspect the same is true
> for NEC-2 and 4.  Does anyone know for sure?
> 
>         Propagation on the low bands is not completely described by
>         current
> models, especially below 2 MHz. 

I'd like to add something we almost always forget. Models are no 
substitute for real world results.

The REASON we have models is the actual system is too complex to 
consider. While the models are certainly very good in some aspects, 
they always miss some things.

For example, our antenna modelling programs assume the earth is a 
perfectly homogeneous media and represent the earth as a single 
boundary layer with certain characteristics. The programs we often 
use do not even consider groundwave, and so show zero radiation at 
zero angles and blend the pattern to that point.

There is no possible way to get everything into the model, and we are 
often not skilled enough or informed enough to know what we are 
missing!

On the other side, they are much closer than we could ever do 
longhand as long as we stay away from limits where they fall apart. 
We have to know the limits of models, just like we have to know the 
limits of measurements. None are perfect, and many are not even 
close. (ie the gain of the antenna is given as 13.65dB but the 
absolute accuracy of the measurement method and instruments is 2dB. 
Why not say "somewhere around 13 dB"?)  

This would especially be true with things heavily dependent on the 
ionosphere, real earth, or any other "soup" of varying 
characteristics, where we might be lucky to get within ten dB. Models 
give you direction, not absolute answers.73, Tom W8JI
W8JI@contesting.com 


<Prev in Thread] Current Thread [Next in Thread>