I n s t r u m e n t s

a n d

M e t h o d s

InstrumentationforIn-situSnowLiquidWaterMeasurements A.Denoth Institute ofExperimentalPhysics, University of Innsbruck, A-6020, Austria

Key Words: Snow, Water, Measurement Techniques, Dielec-

tric Sensors

ABSTRACT

Various methods for measuring the liquid water content of
a natural snow cover are discussed.
An erroranalysisgives suitable measuring quantities:
methods based on the determination of the dielectric func-
tion (permittivity and total losses) at frequencies exceed-
ing 10 MHz allow precise,rapid and nearly non-destruc-
tivemeasurementsand arehighly suitedforfieldappli-
cations.Thespecial designofthedielectricsensorsde-
pends on the actual measuring frequency:flat plate sen-
sors of various geometries have been developed foruse in
the radio frequency regime (20 - 50 MHz),recently tested
monopole-antennas with differentprobe-lengths, operated
in the 100MHz to 3GHz frequency rangehave proofed their
efficiency, and in the microwave X and K bands (8 to 16
GHz) reflection andtransmission measurements havebeen
performed.From broadband measurements of the dielec-
tric functionof snowwith differentstagesof metamorphism
anddifferentliquidwatercontentssimplerelations
between watercontent (volumebasis)snowpermittivity
and densityare deduced. Comparativemeasurements of
the vertical distribution of waterin a natural snow cover
are presented.

INTRODUCTION

The presence ofliquid waterplaysa majorrole insnow
metamorphism,snow mechanics and hydrology and also
otherbranches of snow and soil engineering. In addition,
liquidwaterina snowpackshowsa dominanteffecton
the reflection,absorption and transmission of electroma-
gneticwaves especiallyin the UHF andmicrowaveregimes
used be remote-sensing techniques.
So, there is an important demand for measurementsys-
temsto determine orrecord liquid watercontent in situ

with high accuracy, high resolution in space andtime and
with simplicityin operation. Methods commonly used to
determinethe snowwatercomponentare amongothers
the freezingcalorimetry and electric methods as for exam-
ple TDR, free-spacereflection/ trans-mission measurement
techniques,and relative simple impedance measurement
techniquesusing capacitivesensors. A comparison of error
surfacesof various methodsshows thatthehighestaccuracy
under field conditions can only be achieved by dielectric
measurements. Fig. 1a shows a typical error surfacefor the
freezingcalorimeter (M: mass of snow relative to the mass
of the freezingagent, W(%): volumetric water content, and
dW(%):absoluteerrorinpercent),andatypicalerror
surfacefor electric methodsusingdielectricsensors is given
inFig.1b(P:snowporosity,W(%):volumetricwater
content,and dW(%): absolute errorin percent). Freezing
calorimetry shows anerror surfacewhich dependsstrongly
on the mass of the snow sample used; the influence of the
amountofwaterpresentcan be neglected. Thismethod
requires a very careful operation underfield conditions;
so the measurementrate is limited to 10...15 per hour. The
absolutely lowest measurements errors under field condi-
tions can be achieved by the 'dielectric'methods:dW(%)
<= 0.6%. Comparedto the freezingcalorimetry, thedielectric
methodsarepracticallynon-destructive,asnosamples
have to be physically removed from the snowpack. A more
detailedanalysisof error surfacesof various methodesused
for snow wetness measurementsis given by Colbeck, 1978,
and Denoth, 1994.Because of the simplicityof operation
inthefield,therelatively lowmeasurementerrors,and
thepossibilitytoeasyperformsnowwetness recordings
with microprocessor-assisteddevices,thedielectricmethod
is preferred.
Forpractical in-situ snow wetness measurements,dif-
ferent sensors have been developed and tested recently.

INSTRUMENTATION AND SENSOR DESIGN

129