In article ,
Martin writes:
|
| In the early 1960s civil & military aircraft also supplied weather
| observations in a rather crude form.

I do know a fair amount about such modelling, you know!

The problem about sporadic, crude measurements is that they provide
a VERY bad basis for prediction. Even ignoring the fundamental
inadequacy of such data, writing programs to use it is fiendishly
difficult - MUCH more difficult than writing them to use the data
that comes from a measurement station.

As I said in the first place, up until they got some decent data from
satellites, the primary restriction was the quality of the data.
That lasted until (say) the late 1970s.


Regards,
Nick Maclaren.