Manual classification of fast ice, pack ice, and polynya (open water and thin ice) in Storfjorden from ERS-2 SAR images during winters 1998 to 2001 was used to determine model parameters in a wind-driven polynya width model. Production of ice in the classified areas was then calculated from surface heat balance. The modeled open water area occupied on average 10% of the total area and produced 58% of the total ice (T ice). The volume of brine-enriched shelf water (BSW; V bsw) was estimated to be in the range 0.9?1.1 ? 1012 m3 or 0.06?0.07 Sv (freezing period average) and 0.03?0.04 Sv (annual average). The strength of the northerly wind component seemed to dominate over net heat flux as cause of variability in T ice. Salinity of BSW (S bsw) was found to be primarily governed by frazil ice production, whereas V bsw was mostly determined from T ice and surface salinity (initial and change during winter). Correlation studies of modeled time series of winter mean polynya area and T ice in winters 1970 to 2001 showed that interannual variability could partly be explained by variability in regional ice and ocean conditions, and partly by the strength of the southwesterlies from the North Atlantic or the North Atlantic Oscillation (NAO). Strong southerly winds (high NAO) may give less Arctic ice import and ice production in the western Barents Sea, resulting in a higher surface salinity in Storfjorden in fall. The resulting weak stability in the water column may give favorable conditions for producing large V bsw the following winter. If strong northerly winds over Storfjorden are associated with low NAO, rapid transitions from high to low NAO would give high V bsw and S bsw. The link to NAO seems to vary in time. However, the transition from weak northerly winds in winter 1999 to strong northerly winds in winter 2000 was accompanied by high observed S bsw winter 2000 in agreement with expectations.
Mendeley saves you time finding and organizing research
Choose a citation style from the tabs below