Stable oxygen isotopes from the MOSAIC expedition show vertical and horizontal variability of sea-ice and river water signals in the upper Arctic Ocean during winter

Our aim is to better understand how local winter modification and advected signals from the Siberian Shelf affect the structure of the upper Arctic Ocean along the Transpolar Drift (TPD). Hereto we use stable oxygen isotopes of the water (δ¹⁸O) in combination with salinity to quantify river water contributions (fr) and changes due to sea-ice formation or melting (fi) in the upper ~150m of the water column during the MOSAIC drift. Furthermore, ratios of fi/fr at identical salinities can be used to distinguish waters remnant from the previous summer and those modified locally.

Within the ongoing winter we observed salinification and deepening of the mixed layer (ML) due to sea-ice related brine release together with interleaving waters at the base of the ML and within the main halocline. These interleaving waters with variable sea-ice and river water signals are observed for the first time and have not been observed during summer expeditions before.

The MOSAIC floe drifted in and out of the freshwater-rich part of the TPD and into the Atlantic regime throughout the winter. Despite these strong regime changes the sea-ice related brine content accumulated during the ongoing winter remained visible within the water column. Budgets derived by integration of signals over the upper 100m result in ~1 to 5 m of pure sea-water (34.92 salinity and 0.3‰ δ¹⁸O) removed from the water column for ice formation and are much higher than ice thicknesses of ~0.5 to 2 m observed for the MOSAIC floe. For further evaluation scaling factors have to be considered accounting e.g. for the different densities of ice and water as well as for the lower salinity in the halocline relative to pure sea-water. Therefore, our analysis indicates a lower limit of the advected signal relative to local winter modification within the Arctic Ocean halocline.