Item

Abstract

One of the aims of the "Pilot Study of Evaporation and Precipitation in the Baltic Sea" (PEP in BALTEX) is the improvement of the parameterisation of evaporation over the Baltic Sea in models. The atmospheric regional climate model REMO is used here to simulate evaporation over the Baltic Sea for one year. The turbulent surface fluxes of latent and sensible heat are determined by a bulk formula using prescribed sea surface temperature (SST) values from a coarser grid. Comparison with measurements at four coastal or island sites which can be considered to be marine sites during onshore winds turns out to be problematic: Deviations of the prescribed SST from measured SST are great at certain times and systematic due to seasonal heating or cooling of the water near coasts. A test with realistic SSTs at one site for a 5-days period strongly reduces the heat fluxes. Comparisons with measured fluxes during the periods with strongest evaporation - cold air outbreak with easterly wind in autumn - were not possible due to the specific location of the measuring sites. Three parameterisation schemes are tested in REMO and the effect of e.g. the wind speed dependence of the transfer coefficient C-HN is not as great as expected from tests with artificial data. For future comparisons of regional model results with single- point measurements, measured data in rather homogeneous areas, which are representative for the gridbox area, should be taken. In inhomogeneous areas, mesoscale models may serve as a bridge between the gridbox of the regional model and the measuring point. Application of a fully coupled atmospheric-oceanographic model should improve the SST resolution both in space and time.