[wxqc] Measuring Snow

[Neil Hunt]

Here's an interesting instrument design problem...  My rain gauge is
modified to collect snow for passive solar melting, but now apparently
over-reads by 2-4 *times*.  I'm looking for insight on my modification, or
advice on how to modify it to collect snow for melting, but not over-read
rainfall.  Details below.

I'm operating a remote solar powered weather station reporting via amateur
packet radio (KG6PPD-9), on top a 7200 foot peak about 15 miles west of the
Sierra crest.  The station is based on the Peets U2100 with the pro rain
gauge (parabolic shaped collector with drip formation and counting).  The
solar panels don't have sufficient power to heat the rain gauge, so I
modified it by adding a passive solar melting tube to the gauge: pictures
at: http://tamarackmountain.com/TamarackMountain/Weather/Pictures.html.

The solar melting tube is a piece of black plastic 3-inch drain pipe.  The
Peets pro rain gauge has an 8-square-inch collection area (about 3.19 inch
radius).  The inner diameter of 3-inch drain pipe is 3 inches, and the outer
diameter is about 3.3 inches.  Some careful shaping yields a pipe with a
razor-edge with the same 3.19 inches diameter as the unmodified rain gauge
collector
(http://tamarackmountain.com/TamarackMountain/Weather/Images/DSC_7821.jpg).
Snow that collects in the pipe melts slowly when the sun comes out and the
drips are directed into the raingauge by means of a 3-inch to 2-inch reducer
fitting, and a 3-inch to 3-inch coupler makes a sleeve to prevent snowmelt
or rain from the outside of the collector from tricking into the collection
bucket:
(http://tamarackmountain.com/TamarackMountain/Weather/Images/SnowCollector.j
pg).  The lip of the gauge is about 12 feet above the ground, currently
about 6 feet above the snow.  The site is generally not windy - average
winds of 2-5 mph are common during major rain events.  The design of my
collector was motivated by the CA dept of water resources sensors, which
appear to be a similar vertical tube, albeit of much larger diameter and
height.

3 Teaspoons of water slowly dripped from a small nozzle yields a measured
.11 or .12 inches of rain, as expected.

However, in a large storm, the gauge routinely reads 2-4 *times* more rain
than nearby stations, often showing improbably high figures of 3 or 4 inches
of rain in a day while my neighbors are reading 1.5 inches.  It's possible
that the orography at the site really does extract a lot more water from the
atmosphere, or it is possible that the design of the snow collector somehow
funnels much more rain than the cross-section of the pipe might indicate.
Note that it isn't a case of measuring 1/100 cms instead of 1/100 inches,
since the 3-Teaspoon test comes out correct, and the multiplier between my
measurement and neighbor's varies quite significantly from one event to the
next.

Unfortunately, the site is about 3.5 hours drive away from home, and right
now is about 30 minutes of snow-shoeing from the nearest vehicle access - so
access is strictly limited.  But I was able to visit on one occasion while
it was raining hard, and during that 30 minutes or so, the 2/10ths collected
in a jar on the ground corresponded closely to the measurement from the
instrument.

I made a temporary modification by adding a secondary collection vessel to
the instrument drain.  Snowfall, freezing, and inability to access until the
collection vessel overflowed made this calibration effort useless, but a
side effect was that there was no path for wind to suck through the drip
counter (since the drip tube was underwater).  The gauge still counted
significantly more rain than nearby stations.

My station data can be viewed at
http://findu.com/cgi-bin/wxpage.cgi?call=KG6PPD-9&last=240 (or, if you are
patient, at http://tamarackmountain.com/TamarackMountain/Weather/Index.pl)
while the nearest neighbor is at
http://findu.com/cgi-bin/wxpage.cgi?call=CW4253&last=240.  Note that CW4253
is not solar powered, and accordingly was subject to an extended power
failure this past weekend.  That rain gauge isn't heated either.  We had a
major snow event (4-6 feet) over the weekend, which has barely started
melting through either of our gauges - judging from the readings.  The
nearest CADWR sensor is about 5 miles East at
http://cdec.water.ca.gov/cgi-progs/plotReal?staid=BLD.  This season I've
measured 54 inches plus probably about 5 more still to melt, while the CADWR
measurement is currently showing 34.4 - that's x2 for season total.
However, on Dec 21, 22, 23, I measured a fantastic 9 inches, while BLD shows
only 3.22 inches - which is x8!

Question: Can anyone share details of modification of rain gauges to collect
snow for passive melting?

Question: Is this vertical tube design likely to somehow trap 2-4x as much
rainfall as the same cross-section of rain-gauge?

Question: How likely is it that my site actually gets 3 or 4 inches of rain
in a 24 hour period, while my neighbor 2 miles away only gets 1-2 inches in
the same period?  My hilltop at 7200 feet is on the edge of the Stanislaus
river valley as the first groud over about 6900 feet in the prevailing storm
direction (from the SW); CW4253 is at a similar height, but is behind my
ridge and one other also at 7200 feet.  BLD is also at 7200 feet, but is in
an even more protected location about a mile East of a ridge of about 7300
feet.

Neil/.