Author(s): Jari Haapala; Jonni Lehtiranta; Juha Karvonen; Christian Haas
Linked Author(s):
Keywords: No Keywords
Abstract: The main bottleneck of the development of the sea-ice models is the lack of proper validat ion data. Sea-ice models are typically validated against the observed sea- ice extent, however, the ice thickness field could be biased even the evolut ion of ice extent is correctly modelled. In this paper, we utilize airborne electromagnet ic measurement and satellite SAR der ived ice mot ion for validat ing the operational sea-ice forecasting model o f the Baltic Sea. The model is a mult icategory sea-ice model HELMI, which reso lves ice thickness distr ibut ion, i. e. ice concentrations o f variable thickness categories, redistr ibut ion of ice categories due to deformat ions, thermodynamics of sea-ice, horizontal components of ice velocity and internal stress of the ice pack. The large-scale pattern of the modelled ice thickness is rather similar than the EM derived ice thickness field, but the model underest imate the mean ice thickness. The maximum modelled mean thickness is about 2 meters, but the mean EM ice thickness is 3-4 meters in several locat ions. A deeper analysis shows clearly that the correlat ion between the modelled mean ice thickness and mean EM-thickness is rather weak and it indicate direct ly that the production of the ridged ice is underest imated in the model simulations. Since the production of the ridged ice is a result of the spatial gradients of the ice velocity fields, possible reasons of the underest imat ion of the mean ice thickness are due to i) neg lect of shear deformat ions, ii) rheolog ical parameters of the model yield too stiff ice mot ion or iii) the wind stress used for the model simulat ion is too low.
Year: 2010