Validation of HIRLAM boundary-layer structures over the Baltic Sea

The boundary-layer structures of the operational atmospheric model HIRLAM were validated over the Baltic Sea on the basis of rawinsonde soundings and surface-layer observations during r/v Aranda expeditions. The validation was made for two regions in 1999: a coastal sea ice zone in March and the Baltic Proper in October. In March, HIRLAM wind analyses and six-hour forecasts were very good. The main discrepancies were related to the surface and 2-m temperatures: in cold nights the inversions were too weak and delayed in HIRLAM. Experiments applying a two-dimensional mesoscale model suggested that HIRLAM results could be improved by updating the values of surface albedo and the parameters of the force-restore surface temperature scheme on the basis of the snow age and temperature. In October, the temperature profiles were accurate within 0.5 K, on average, but the boundary layer was too moist in HIRLAM. The wind speed in the analyses and six-hour forecasts was accurate within 1 m s(-1), and errors in the sea surface temperature had a strong effect on the turbulent surface fluxes.