MASS2ANT Surface (10 meters height ) wind field Dataset (Downscaling CESM2 @5.5km over Dronning Maud Land, Antarctica, 1850 - 2014)

We provide in this dataset maps at 5.5 km resolution of the  yearly mean of surface 10m wind components over a region encompassing Dronning Maud Land (Antarctica) from 1850 to 2014. We used a statistical method to derive fine resolution maps from GCM runs (CESM2, 10 runs). In the method, we searched for analogs in a database we constructed from the association between re-analyses large-scale meteorological fields (ERA5 and ERA-Interim) and RCM daily accumulated snowfall (RACMO2.3p5.5). RACMO2.3p5.5 data are available freely on request (https://www.projects.science.uu.nl/iceclimate/models/antarctica.php). CESM2 CMIP6 runs are also freely available (https://esgf-node.llnl.gov/search/cmip6/). This dataset is an extension of the snowfall dataset, for which a complete description of the algorithm and performance used for snowfall is in: Ghilain N., Vannitsem S., Dalaiden Q., Goosse H., De Cruz L., Wei W., Large ensemble of downscaled historical daily snowfall from an earth system model to 5.5 km resolution over Dronning Maud Land, Antarctica, Earth Syst. Sci. Data, 14, 1901–1916, 2022.

The MASS2ANT wind fields dataset is used for further analysis in: Cavitte M., Goosse H., Dalaiden Q. and Ghilain N., Brief Communication: annual large-scale atmospheric circulation reconstructed from a data assimilation framework cannot explain local East Antarctic ice rises’ surface mass balance records, accepted for publication in The Cryosphere (10.5194/egusphere-2024-3140)

The MASS2ANT wind components dataset is composed of the mean annual estimations of wind components over Dronning Maud Land, based on the daily time series for the total period for all grid points of the domain, excluding a band of 8 pixels  at the boundary. The method to downscale re-uses the associated principal components time series and Empirical Orthogonal Functions (EOF) (stored on zenodo in associated datasets) offering the possibility to analyze the synoptic weather patterns associated to snowfall over the ice sheet and the Principal Component weights (PCs) time series from the re-analysis in case one wants to extend or improve the database. Realistic weather patterns can be recomposed in associating (product of matrices) the PCs with the EOFs.

Here, we provide the annual averaged wind components time series resulting from the downscaling of the 10 members of CESM2 (index r1 to r10 in the file name), using ERA5 and RACMO2.3p5.5 for training.