Model results for hydrography are here compared with observations from the Hirtshals (Denmark) - Torungen (Norway) transect. Model results were stored at the measuring depths of the transect data. Generally, errors are reduced with more advanced representations of the relevant processes. The results for the fresh water height show that the modelled salinities are too low on the Danish side and in the central Skagerrak, whereas the salinities are too high on the Norwegian side.

Validation of various representations of river run-off and Baltic fluxes in MI-POM simulations, based on observations from the Hirtshals (Denmark; to the left) - Torungen (Norway; to the right) transect. Model results and observations are from 1998. Results from the following experiments were considered:
• Control expt.: Climatological run-off from all rivers, and climatological fluxes of Baltic outflow to the Kattegat
• Expt. 1: Climatological run-off replaced by the most realistic run-off data available; still using climatological fluxes for Baltic outflow into the Kattegat
• Expt. 2: Realistic river run-off as in Expt. 1, but incorporating a simple analytic "Baltic reservoir" sub-basin model for volume fluxes, with outflow salinities set to 18 PSU)
• Expt. 3: Realistic river run-off as in Expt. 1, but incorporating results from a storm surge model for volume fluxes, with outflow salinities set to 18 PSU)

Results for the fresh water height, based on the same observations and model results as the top panel. The fresh water height is computed as the vertical integral of
    max[S₀ - S(z)]/S₀
where S₀ is the salinity of the water masses that are unaffected by the regional fresh water fluxes (here, S₀ was set to 35 PSU).