Hence, the model suggests a more complicated interaction of the frontal processes with the cavity circulation, and a full investigation of this transient response to the time-varying forcing check details will need attention in future work. The simulated melting beneath shallower parts of the FIS appears to be determined by the combined effect of sub-ice shelf currents and hydrography. For all hydrographic scenarios, stronger winds increase the shallow melting (P3 in Fig. 10), because a more energetic upper ocean circulation (Fig. 9(d)) enhances

the exchange of ISW with warmer ambient water beneath the ice, and stronger currents also increase the parameterized mixing at the ice shelf/ocean boundary. ZD1839 nmr Accordingly, the

experiments with stronger winds show more surface water beneath the ice, indicated by the salinity contours on top of the temperature shading in Fig. 5(c)–(e), and the more frequent occurrence of buoyant water in the θθ-S histograms in Fig. 6(d)–(f). The surface layer speeds in Fig. 9(d) also show stronger currents for the weak wind experiments in the ANN- and SUM-scenarios that are not consistent with this theory. However, this is likely an internal melting feedback, where strong deep melting produces highly buoyant plumes that rise along the ice base and dominate the shallow flow field in these simulations. The varying hydrographic conditions are found to have two opposite effects on the shallow melting response in the different experiments. One effect is that ASW increases the melt rates by replacing the cold ISW with warmer waters near the ice base as described in Section 4.3 (P5 in Fig. 10). The opposing effect is that larger amounts of buoyant surface water in the model reduce the shallow melting by weakening the near-surface currents (Fig. 9(d)), as demonstrated by comparing the circulation between the ANN-100 and the WIN-100

experiment in Fig. 8. In order to separate the dynamic control of the ASW (P4 in Fig. 10) from its role as an additional heat source, an additional model experiment was conducted, in which the hydrographic forcing uses the constant summer scenario to restore the salinity, but applies the constant winter scenario with all waters above the thermocline Flavopiridol (Alvocidib) at surface freezing-point for restoring the temperatures. The result is an upper-ocean circulation that is as weak as in the constant summer situation, and shallow melt rates that are even weaker than in the constant winter scenario. This shows that the density of ASW, being mainly controlled by salinity, can counteract the melting increase caused by warmer temperatures. A more detailed analysis (not shown) reveals that the weaker upper-ocean currents not only decrease the friction velocity in the applied basal melting parameterization, but also reduce the mixing of the ISW beneath the ice base with the (warmer) ambient water in the cavity.