4). In addition, it is unclear how the temporal and spatial scales of the time-varying wind field would affect the circulation in the limited model domain, possibly causing circulation artifacts due to interference at the periodic boundary. Another simplification that is required to ensure consistency at the periodic model boundaries is the omission of tidal forcing. Propagating tidal waves would interfere with their images at the cyclic model boundary. Also the successive superposition of tides in separate non-cyclic model runs was found to strongly alter the mean circulation, Etoposide leading to the development of

circulation artifacts (Abrahamsen, 2012). However, tidal currents in the Eastern Weddell Sea region are generally rather weak (Padman

et al., 2002). A discussion on how tides, sea ice and time-varying winds may alter our results will be given in Section 6.3. In addition to the semi-idealized ANN-100 experiment, we study the melting response to different climatic conditions by systematically varying the idealized model forcing. The role of easterly winds for the momentum balance of the ASF current is explored by varying the magnitude of the wind stress by a constant factor, here denoted by percentages with “100” indicating the RACMO2 average. A strong wind forcing (denoted “130”) MDV3100 nmr with 130% of the average surface stress, as well as four weak wind forcing forcings (30, 40, 60 and 70), are applied. This range was chosen to highlight the two possible states of melting that are revealed by our simulations. The effect of the ASW formation is investigated by using different hydrographic conditions for the water mass restoring at the surface and the lateral boundaries. In addition to the time-varying annual cycle scenario described above (denoted

ANN) a constant summer (SUM) and a constant winter (WIN) scenario are used for the hydrographic nudging. In the constant winter scenario, no ASW is present and a homogeneous layer of ESW with temperatures at the surface freezing point occupies the water column above the thermocline. The constant summer scenario is defined by the mid-April climatology indicated by the dashed line in Fig. 3(c), when the distribution of ASW extends deepest throughout the water column. Combining nearly the different wind and hydrographic forcings, 18 different experiments, as denoted in Table 1, were preformed. Each experiment starts from an initialization state at equilibrium, produced by a 10 year spin-up with the constant winter (WIN-100) forcing applied and the model being initialized with temperatures at the surface freezing point, a horizontally uniform salinity profile, and zero velocities. The initialization state reproduces a fully developed ASF mesoscale eddy field, as illustrated by the snapshot of relative vorticity in Fig. 2(b).