"Hugh Newbury" wrote:

Martin, It occurs to me that maybe the answer might be to restrict
the outlet from the AWS to its tipping bucket. Small as it is,
possibly it's still too large to cope with a deluge. The weight of
the rain in the bowl would force the water out more quickly.
Certainly my experiments today seem to suggest this. I'll think
about it for a day or two.

.... as John suggested (elsewhere in the thread) I would be reluctant
to restrict the flow - it might (almost certainly will) lead to bigger
problems.

Indeed, I'd be reluctant to do anything to *physically* alter the
collector/gauge until I was absolutely sure that the difference wasn't
'genuine' in the sense that there isn't another (external) factor
accounting for the results.

Looking at your data, you get some very close figures when the rain is
'steady', but the difference multiplies alarmingly when the rain is
intense and accompanied by strong winds (which as others have stated
seems to have been a feature of this summer), which would steer me
towards an exposure problem, which the loss due to (attempted) rapid
cycling of the sample accentuates.

I know this is going to sound 'clunky', but you might want to consider
mounting a passive collector adjacent to the AWS gauge (i.e. same
height/not more than 30cm apart) - even a jam jar and a plastic funnel
would do, just to see if it collects roughly the same amount of rain
at that location. It's not ideal as one will interfere with the other
of course, but I would hope that the difference wouldn't be more than
10%; at the moment, during some of your events you are getting
differences of 30% or so.

Martin.


--
Martin Rowley
West Moors, East Dorset (UK): 17m (56ft) amsl
Lat: 50.82N Long: 01.88W
NGR: SU 082 023