My thanks to all who responded. Individual contributions follow. Note that
the scale is intended as a measure of the likely subjective impact of
un-brought weather, not driving conditions. Readers are welcome to let me
know their views, but the 'public' side of the discussion ends here.

Regards
Miland Joshi 

*** Summary of contributions***

I love this. Nice to know there are other pockets of sanity in the universe.

BTW we have
rain 1 trying to be 2 
cold 1
wind 1
So we are miserable at the level 3 also (though one could whine and claim a
4)


Yours


Steve Beaney

***
good idea! - there is a lot of research on the effects of 
weather on health, check out the International Journal of 
Biometeorology, for example - however, whilst myself and 
others have found statistical associations between 
temperature and many types of illness, from flu to heart 
attackes and strokes, and even between air pressure and 
circulatory diseases, not much is known about the 
psychological effects of 'miserable weather' which may then 
lead to increased psychosomatic illness. The Met Office has 
daily measurements of all sorts of weather indicators for 
120 stations in England and Wales, so data is no problem, 
but the biological pathways that link weather to health are 
very poorly understood - for example, why do people get 
colds in winter? there are theories, but no-one is really 
sure....

Ian Langford, CSERGE, UEA

***

There is an existing scale for each of these. Wind - the Beaufort 
Scale (but see http://www.crh.noaa.gov/lot/webpage/beaufort/) ; 
temperature - the Celsius, Farenheit or Kelvin Scale and rain - mm/hr 
(or other combination of height within a flask and time). It may be 
sensible to produce an ordinal scale with a suitable number of points 
out of the two metric continuous scales, the Beaufort scale is already 
ordinal. Also, if degrees Celsius is used one would need to be careful 
about negative temperatures. The numerical values on each of the three 
ordinal scales could reflect the relative weight of each facet of the 
weather in terms of its 'badness'. For example, if all scales are 
12-point, like the Beaufort scale, but rain is deemed to be twice as 
bad as wind, we could use 1 to 12 (or maybe we should rescale it 0 to 
11) for wind and 2 to 24 ( or 0 to 22) for rain. The three scales could
then be compounded to form a 'bad weather score'. One might also wish 
to take account of duration of a spell of weather. I find a whole day's
drizzle much worse than a transient heavy downpour. There is a further 
issue regarding temperature which is particularly exercising me. My 
ideal weather is bright clear blue skies, a nice light breeze and a 
temperature of about 10 degrees Celsius (I'm not a sun worshipper, you 
see). I would therefore suggest that above a certain temperature the 
temperature becomes 'bad' again, what this threshold should be, or 
whether there should be a threshold at all would obviously be a
subjective decision, about which there would no doubt be some heated 
(pardon the pun) debate.

Andy Scally


***

I can't suggest any weights, but I'm sure there'll be allstat 
subscribers in other parts of the world who'll be fascinated by the 
idea of it raining cats and dogs here.  Language can play some 
strange tricks - yma yng Nghymru mae hi'n bwrw hen wragedd a ffin - 
here in Wales *it* doesn't rain cats and dogs, but *she* rains old 
ladies with sticks!

Robert Newcombe.

***

Clearly the meteorologists, at least here in the US, have a way of weighting
wind
against temperature, as they quote "wind chill factors": the apparent
temperature
given the wind speed. But maybe the inconvenience of wind (umbrellas blowing
inside-out etc.) exceeds merely its cooling effect.

Rachel Pearce

***

What you are describing is close to the 'wind chill index' used in the USA
in
winter to exaggerate the climate severity (and some say, hide the
approaching
global warming :)  ).

Go for it!  Make up a new scale.  Include rains of cats & dogs!  Maybe
raccoons
(4 of which were born under my deck this spring - does that count?)  You'll
need to make gradations for wee Scots mists.  Don't forget snow - in
northern
WI, USA this past weekend, they got near a foot (whoops - almost 30 cm).

If you do it right, Leichester may regain title to the most raw weather in
GB!   But seriously, I'll bet there would be a good correlation between that
and some specific illnesses.

Cheers,

Jay Warner

***

In the creating of the ordinal scale measuring faul weather, you must define

the goal of the scale. Do wish to warn elderly people not go outside if the 
score is higher than say, 4? Or should people store their car in a garage 
whenever the score is higher than 5? The purpose of the scale will determine

its definition (and weights corresponding to constituent scales, e.g. 
temperature). Practical experience however, suggests that such scales cannot

be made foolproof.

Regards,
Jarl Kampen


%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%