In article , Andy Wade
writes
tony sayer wrote:

They cannot take into account local obstructions like buildings and
other obstructions other than adding in an "urban clutter" factor which
can be very variable!...

Oh yes they can. The UK planning model for DTT uses real clutter data -
see for example:

http://www.bbc.co.uk/rd/pubs/whp/whp048.shtml

"The paper outlines the approach for predicting received field strength
with particular discussion of profile extraction, radial prediction and
the use of clutter data to take into account the effect of buildings and
trees."

Yes, but we were talking about the postcode predictor were we not?, and
that is inherently inaccurate because a postcode can cover quite a large
area. So much so that the elevation could vary by a very significant
amount in the given area, so you have to take a mean value hence the
inaccuracies of this method.

In practice however its just an indication and nothing more, and perhaps
better then the more conventional maps that are used such as coverage
maps from the BBC for example.....
--
Tony Sayer