"Philip Eden" philipATweatherHYPHENukDOTcom wrote in message
...
This subject comes up regularly, and the received wisdom
is that AWSs provide tremendous detail about the character
of rainfall events, but - especially for those guages in an
elevated position - catch less than a standard manually-read
gauge (127mm diameter copper gauge with its rim 30cm
above the ground.)

The Eden mantra is: "Always install a standard gauge
alongside your AWS".

Here are some comparisons (Manual/AWS):
At Luton the two sites are 1.2km distant so there will
occasionally be some geographical difference between
them, but the land is pretty flat and the difference in
altitude is only 2m. The AWS gauge is 1.6m above
the ground:

July 2007 1.04 (significant geog. difference on one day)
August 1.12
Septmbr 1.24 (significant geog. difference on one day)
October 1.14
Novmbr 1.15
Decmbr 1.08
Jan 2008 1.13
Mean 1.13

At Chesham, the two gauges are on the same site, but
the AWS gauge is 1.6m above the ground:
Oct 7 - Nov 6 1.04
Nov 7 - Dec 2 1.07
Dec 3 - Dec 17 1.07
Dec 18- Jan 4 1.06
Jan 5-28 1.05
Mean 1.06

Philip Eden

This interests me, since I have just installed a 5" copper gauge to back up
my Davis VP2 AWS. So far (only five rain days this month and allowing for
'throw back' of 09Z manual readings) the 5" is measuring about 15% higher.
Obviously such a small sample is pretty meaningless, and could be swamped by
the different 24-hour measuring spans, but I wondered what other observers
are seeing, and how much of the under-reading by the AWS is caused by lost
tips? I haven't calibrated the AWS by pouring water into it, by the way -
perhaps I should.

Steve P
Acton Bridge 38m
http://www.pardoes.com/meteo/weather.htm