Snow modelling using Surfex with the Crocus snow scheme for Norway Dagrun Vikhamar-Schuler 1 and Karsten Müller 2 1 MET.NO, 2 NVE
Data sources Weather forecasts Weather stations Gridded 1 km maps: snow depth, temp, prec. Snow observations Physical snow models seNorge.no
CROCUS snow scheme: -Up to 50 layers -Temp -Grain size, type -Liquid Water cont. -Density -.. SURFEX: land-surface model from HARMONIE atm. model. Ref: Eric Brun and Samuel Morin, MeteoFrance Date Snow depth Snow temperature
1D model run for weather stations in Norway
Driving data? Short- and longwave radiation Precipitation (+prec. phase) Air temperature Wind dir. + speed Surface pressure Air humidity Too few parameters are observed at the weather stations to drive SURFEX. Can forecasts replace observations without reducing the quality of the modeled snow depth?
Grotli (872 masl): forcing data: observations except …: Observed Snow Depth Modelled Snow Depth UM: Radiation UM: Radiation, Surface pressure UM: Radiation, Wind UM: Radiation, Air humidity UM: Radiation, Temperature UM: Radiation, Precipitation
Sensitivity study For all Stations (Grotli, Filefjell, Beitostølen, Mannen, NGI- fonnbu): BEST results using OBSERVATIONS of precipitation and temperature. Other parameters seems ok to derive from MetUM (NWP) forecasts.
We have evaluated the snow depth. What about the snow pack layers?
Liquid water content (%) Filefjell winter Snow density(kg/m3) 1 Sept April 2012
Snow Temperatures FILEFJELL Weather Station Snow temperatures are measured hourly 0, 5, 15, 30 og 55 cm snow height ºC Ground surface temperature
Summary CROCUS reproduces well the observed snow depth using observations. Results are most sensitive to precipitation and temperature Other parameters ok to derive from NWP data Other parameters ok to derive from NWP data Ongoing work: comparing driving data: NWP models (Harmonie 2.5 km) NWP models (Harmonie 2.5 km) Observations Observations Future: 2D simulations