Floating ice shelves regulate Antarctic ice sheet mass loss by buttressing land ice discharge toward the ocean. Next to basal melting, iceberg calving following the propagation of rifts has the potential to reduce this buttressing effect. However, rift propagation is largely unpredictable and generally not resolved in coarse climate models. The sub-shelf ocean circulation is believed to play an important role in rift propagation as it is closely related to ice shelf melting and freezing. Previous work suggests that the sub-shelf circulation of an intact ice shelf is altered by the presence of km-wide basal channels. As it could impact freezing and melting, the potential effect of basal channels on a rifted Antarctic ice shelf should be explored.

In this study, the Massachusetts Institute of Technology general circulation model was applied to explore for the first time the potential effect of basal channels on the melting and freezing of a rifted Antarctic ice shelf. To this end, four simulations were performed with a high-resolution ocean model for the domain of an ice shelf cavity with idealized boundary conditions. These simulations correspond to a cavity with melt channels, a prominent rift close to the ice shelf front, both, and none of them. The effects of channels, a rift or the combination of both on melt, freezing and ocean circulation in the cavity were assessed through a comparison of these four simulations. Following previous research, results show that basal channels decrease ice shelf basal melting. We found that the addition of only a rift does not change the melt intensity or pattern. In addition, it was found that basal channels increase the freezing inside a rift. A sub-shelf boundary current on the Coriolis favoured side of the domain without channels is reformed to a clockwise circulation in each channel, resulting in an adjusted flow pattern inside the rift from one single large clockwise return flow to a smaller one behind each channel. Hence, buoyant cold shelf meltwater does not only enter the rift in the boundary current but after every topographic incision, and the thermal forcing is increased. Furthermore, the multiple return flow pattern enlarges the average frictional velocity inside the rift, which is positively related to the freezing rate intensity. In an offline calculation, it was found that the contribution of the thermal forcing to the total freezing amount is approximately three times larger than the friction velocity.

From our simulations, we conclude that the presence of basal channels in a rifted ice shelf decreases basal melting at the grounding line and increases freezing inside rifts. Previous work on rift propagation suggests that these results imply that marine ice accretion inside the rift could increase, and fracture propagation could be reduced. Given these connections between ice shelf processes, this study stresses the importance of including basal channels and rifts in ice shelf cavity models to robustly reproduce Antarctic sub-shelf circulation and basal melt.

The amount of suspended sediment spill released during dredging activities in the Fehmarnbelt strait is monitored along the excavated trench to prevent negative effects on local ecology. Stratification due to the interaction of saline water from the North Sea and fresh water from the Baltic Sea forces bi-directional flow of the layer above and below the density gradient, causing bi-directional spread of the dredging plume perpendicular to the trench as well. During such an event, monitoring on both sides of the trench is required. By mapping out the spatial and temporal behaviour of stratification, a prediction can be given on where and when measuring on both sides of the trench will be needed to include the baroclinic effect. Figures created with monitoring data show that the density profile is influenced by the salinity, rather than the temperature. Therefore, stratification predominantly depends on the in and outflow of water with respect to the Baltic Sea. The figures also show a larger density gradient during out than inflow. It can also be observed that where the water depth is restricted to 10 meters, wind and bottom friction mix the entire water column. Therefore, stratification occurs predominantly during outflow in sections deeper than 10 meters, indicating the need for monitoring the bi-directional plume spread during such circumstances. Whether stratification occurs during inflow in sections deeper than 10 meters likely depends on the duration and strength
of the wind forcing and the initial strength of the density gradient prior to the inflow event. Further analysis should be done to confirm this. Signs of Ekman transport, return flows and the deflection of the currents towards deeper water can also be observed in the measurement figures. Since these processes affect the plume spread direction, additional research can be done on the behaviour of the current direction in the Fehmarnbelt.