Weather Forecast
Pour nous contacter
an image
Université de Liège

Faculté des Sciences

Département de Géographie

Laboratoire de Climatologie et Topoclimatologie

Quartier Village 4

Clos Mercator 3, B11

4000 Liège


Phone: +32 (0)4 366-52-21

The Laboratory » Team » Dr. Xavier FETTWEIS » SMBMIP


Welcome to the SMBMIP intercomparison exercise!

(last update 11/10/2018)


This MIP exercise, requested by ISMASS and ISMIP6 at Polar 2018 (Davos), aims to evaluate all estimations of SMB, coming from RCM, PDD and GCM over the Greenland ice sheet in the goal of forcing ice sheet models afterwards. The comparison will focus as a first step over the current climate (1980-2017), at a resolution of 1km and at a monthly time scale. The PROMICE SMB data base, ice cores and satellite derived melt extent as used in Fettweis et al. (2017) will be notably used to evaluate the different estimations at monthly time scale.


Deadline: 20 October 2018. The first results should be presented at AGU 2018 in December.

Period: 1979-2017
For RCM and PDD: ERA-Interim as forcing.
For GCM: Historical scenario over current climate or an ERA-Interim driven AMIP scenario.

Temporal resolution: monthly

Spatial resolution: 1 km

Grid: In the aim of unifying the used grid, the best is to provide your outputs on the 1681 x2881 EPSG:3413 ISMIP6 1km grid.  An example of this grid is provided here. If you provide your outputs on another grid, like the 1496 x 2700 1km grid used by Noel et al. (2016), I can perform myself the interpolation but I need lat/lon of each of your pixel in this case.

Fixed fields: latitude, longitude, topography used natively by the model, topography used in the downscaling technique, ice sheet mask used natively by the model, ice sheet mask used in the downscaling technique. The comparison will be performed on the common ice sheet mask of all models.

SMB components: SMB, snowfall, rainfall, meltwater, run-off, total water flux (i.e. sublimation+evaporation(+wind snow erosion)-deposition-condensation) in mmWE/month. If it is applicable, both raw (i.e. just interpolated to the 1km grid) and downscaled/corrected outputs can be provided. See the MARv3.9 example where both SMB and SMBcorr are provided. 

SEB components: monthly average of incoming shortwave, incoming longwave, sensible heat flux, latent heat flux, albedo + (near) surface temperature. 

Storage of output:

User: smbmip
Password: (send an email to xavier.fettweis @

Format: NetCDF. One file by year with all the requested variables is the best.

Example: MARv3.9 (run at 7.5 km and downscaled to 1km) outputs for 2017 (but on the 1496 x 2700 grid !).

ContactXavier Fettweis