I n s t r u m e n t s

a n d

M e t h o d s

mean wind velocity of 10 m/s. On the graph(figure 5), the
behaviourofeachsensorisdifferent,dependingofits
position along the slope. For example,the sensor1 seems
to measure a smaller thickness of snow accumulation than
the others.However,the snow rate of erosion is approxi-
mately thesame in the4 cases,exceptfor thesensor 2 which
display alternately erosion and accumulation phases.

Correlation with weather parameters

During the same period, if the standard-deviation of the 4
hourly differencesof profileur datais compared separately
with precipitation or wind speed, it appears that the snow
surface evolution cannot be explained by a single param-
eter (figure 7).
In ordertounderstand the mechanism of snow trans-
port by wind otherparameters are needed: snow particle
characteristics and grain cohesion.
Afteran additional winter season of data, we expect to
have interesting snow transport records (it was not really
thecaseoflastwinterbecauseoftheprevailingnice
weather!).So,Proteon (Guyomarc'h and Mérindol, 1995)
will be used to analyse a snow pack simulated by Crocus.
The correlation between the snow transport index calcu-
latedbyProteonandthehourlyevolutionofthesnow
"profileur" could be interesting.

CONCLUSIONS

The purpose ofthis paperwas toshow the feasibility of
automatic snow-depth measurements along a slope and to
getabetterideaofthecorrelationwithmeteorological
parameters on the one handand with snow grain features
on the other hand.
Thefieldmeasurementsaremadeunderconditions
whicharesometimesverybad.Nevertheless,these pre-
liminary observations allow us to make some concluding
remarks:thisdevice seems,in afirstapproximation,to
workregularlyevenunderheavysnowfallsandstrong
wind,thisequipmentshouldpermitustopreciselyde-
scribe the evolution of a snow pack whichundergoes wind
effect. For this reason, we want to have one more season of
field measurement records.
The final aimisto take into account the behaviourof
snow under wind effect andto improve snow transport by
wind and snow pack evolution modelling software.

ACKNOWLEDGEMENTS

This research has been feasible with the financial support
of the "Pôle Grenobloisd'Etude et de Recherche pourla
PréventiondesRisquesNaturels"(GrenobleCentrefor
Study and Research on Natural Hazard Prevention).This
study has alsorequired the co-operation between ourre-
search centre and the "Alpe d'Huez" ski resort.Fortheir
commitment tothe project,we thank the "SATA" (Snow
Safety Service of Alpe d'Huez). Of course, therewould not
be any results without the laboratory and field work of the
CEN's electronic team.

134

REFERENCES

CastelleT.,1994Mécanismes et simulations dutransport de
la neigeparle vent enmontagne:Applicationau site experi-
mental duCol duLacBlanc.Ph.D.E.P.F.Lausanne.(Suisse)
297 p.

Guyomarc'h G., Mérindol L., Castelle T., Sivardière F., Buisson
L.1994BlowingSnow andSlabAvalanches.Proceedings of
94'ISSW Snowbird-Ut (USA).

Guyo marc'h G., Mérindol L. 1995 Protéon: vers une prévision
locale du transport de neige par le vent. Colloque Anena Cha-
monix pp 97-102.

Lecorps D.,Sudul M.198910years of snow depthmeasure-
mentsproceedingsofInternationalWorkshoponPrecipita-
tion Measurements St Moritz (Switzerland).

RoccaV.1996Traitementstatistiquedesdonnéesd'un
profileurdu manteau neigeux.Rapport DUT2^èmeannée.

Sivardière F., Castelle T., Guyomarc'h G., Mérindol L., Buisson
L.1994Functioningoftheavalanchestartingzoneswhich
undergo snow-transport by wind: field observations and com-
putermodelling.SurveysinGeophysics.EGS,25-29April
1994.Grenoble (France).