The thread was getting a bit long, so I will add my penny's worth here.

Part of the problem with recording amounts for frost/dew etc is that much,
probably nearly all, of any water that finds its way into the gauge bottle
on a standard 5 inch copper gauge as a result of dew/frost is in fact caused
by condensation on the gauge surface, some on the inside. If the inside of
the gauge is kept dry, this will reduce the effect considerably. For some
years I have been keeping three dedicated plastic bottles for collecting the
pptn in the standard gauge. Each morning I take a dry bottle to the site and
exchange it for the one that has been in the gauge for the past 24 hours. I
measure the weight of the precip while still sealed in the bottle, before
pouring it into a standard measuring jar. An accurate weighing scales will
give results to 0.01 of a mm. Usually, (but not without exception), and the
past three days are an example, the bottle in the 5 inch gauge is bone dry,
even if the TSR and the TBR record 0.1 mm, showing that even a thick hoar,
such as we had on the morning of the 13th, will not necessarily produce any
drips in the 5 inch gauge.
In my view it is not possible to differentiate from natural deposits of
dew/frost on the exterior of a gauge, and condensation that occurs on the
surface as a result of a rise in temperature. Using my dry bottle each day
system, I almost never see any wetness on the interior surfaces of the
gauge, in contrast to earlier times when I relied on a glass collecting
bottle and measurement on site. It is impossible to get all the water from
the collecting bottle into the measuring jar, and a wet bottle will thus
always be returned to the gauge to provide water vapour to the interior
space that will often end up as condensation there.

--
Bernard Burton
Wokingham, Berkshire, UK.

Satellite images at:
www.woksat.info/wwp.html
or
www.btinternet.com/~wokingham.weather/wwp.html