Hubert Lamb does this in Historic Storms of the
North Sea, British Isles and Northwest Europe, published
by CUP, 1991. He discusses different kinds of grading,
from the objective
windspeed^3 * Area affected * Duration

(choosing windspeed^3 rather than windspeed^2 as
representing wind power rather than simply wind force)

an estimation of total damage to the landscape,
or the number of deaths and injuries
or insurance losses

which are each open to interpretation.

I've been toying with the idea of extending this
to all major weather events ... but the problems of
comparing say a windstorm with a killer smog
are virtually insurmountable.

Philip,

I think you've hit upon the major problem here, any classification is likely
to be very subjective.

The largest sea, and the most damage to the Penzance/Newlyn/Mousehole sea
defences for 40 years was done on 27th October 2004.
The depression itself was nothing really special, and if you lived 2 miles
inland you would have hardly noticed it. It was the track and speed of
movement of the depression, which generated a massive sea from an unusual
direction, combined with an exceptionally high tide which created the
problem. As far as sea conditions are concerned, the reverse of what Richard
Dixon suggests
'that the rapid spin-up systems tend to be the most damaging' that is
normally true.
It's hard to see what classification system would treat it as the major
event it was locally.

Even if you had a 'major coastal event' classification, there would be
difficulties in comparing say the N Cornwall coast where a swell of 10 feet
is not unusual (it was 10-14 feet a couple of days ago at Gwenver north of
Land's End), with a south coast bay where it certainly would be.

Graham