Boundary-Layer Meteorology

, Volume 117, Issue 2, pp 337–381

‘Modelling the Arctic Boundary Layer: An Evaluation of Six Arcmip Regional-Scale Models using Data from the Sheba Project’


    • Department of MeteorologyStockholm University
  • Mark Žagar
    • Department of MeteorologyStockholm University
  • Gunilla Svensson
    • Department of MeteorologyStockholm University
  • John J. Cassano
    • Cooperative Institute for Research in the Environmental Sciences and Program in Atmospheric and Oceanic SciencesUniversity of Colorado
  • Susanne Pfeifer
    • Max-Planck Institute for Meteorology
  • Annette Rinke
    • Swedish Meteorological and Hydrological Institute
  • Klaus Wyser
    • Alfred Wegener Institute for Polar and Marine Research
  • Klaus Dethloff
    • Swedish Meteorological and Hydrological Institute
  • Colin Jones
    • Alfred Wegener Institute for Polar and Marine Research
  • Tido Semmler
    • Max-Planck Institute for Meteorology
  • Michael Shaw
    • Swedish Meteorological and Hydrological Institute

DOI: 10.1007/s10546-004-7954-z

Cite this article as:
Tjernström, M., Žagar, M., Svensson, G. et al. Boundary-Layer Meteorol (2005) 117: 337. doi:10.1007/s10546-004-7954-z


A primary climate change signal in the central Arctic is the melting of sea ice. This is dependent on the interplay between the atmosphere and the sea ice, which is critically dependent on the exchange of momentum, heat and moisture at the surface. In assessing the realism of climate change scenarios it is vital to know the quality by which these exchanges are modelled in climate simulations. Six state-of-the-art regional-climate models are run for one year in the western Arctic, on a common domain that encompasses the Surface Heat Budget of the Arctic Ocean (SHEBA) experiment ice-drift track. Surface variables, surface fluxes and the vertical structure of the lower troposphere are evaluated using data from the SHEBA experiment. All the models are driven by the same lateral boundary conditions, sea-ice fraction and sea and sea-ice surface temperatures. Surface pressure, near-surface air temperature, specific humidity and wind speed agree well with observations, with a falling degree of accuracy in that order. Wind speeds have systematic biases in some models, by as much as a few metres per second. The surface radiation fluxes are also surprisingly accurate, given the complexity of the problem. The turbulent momentum flux is acceptable, on average, in most models, but the turbulent heat fluxes are, however, mostly unreliable. Their correlation with observed fluxes is, in principle, insignificant, and they accumulate over a year to values an order of magnitude larger than observed. Typical instantaneous errors are easily of the same order of magnitude as the observed net atmospheric heat flux. In the light of the sensitivity of the atmosphere–ice interaction to errors in these fluxes, the ice-melt in climate change scenarios must be viewed with considerable caution.


Arctic climateClimateClimate modelNumerical modelling
Download to read the full article text

Copyright information

© Springer 2005