2. Sufficiently small pressure gradient errors are commonly believed to alter the solution by linearly superimposing a geometry-dependent spurious component to the background flow. To assure that these effects are minimized, several tests with various realistically BAY 80-6946 in vivo stratified but horizontally uniform profiles of temperature and salinity were performed. In these test cases, which ideally should produce an equilibrium state that is fully at rest, the maximum velocities occur near the ice front, but remain small (below 2 cm s−1) relative to

the typical 5–50 cm s−1 currents occurring in the full simulation. In order to estimate the influence of different oceanic processes on basal melt rates, a set of semi-idealized model forcings is derived from the data presented in Section 2. The forcing which most realistically represents the FIS present-day conditions, referred to as experiment “ANN-100” hereafter, assumes a quasi-steady annual cycle of the coastal circulation

selleck chemical and can be described as follows. To reproduce realistic water masses in the model interior, temperature and salinity at the eastern (inflow) model boundary are nudged to the time-varying climatological ASF section described in Section 2.2. The nudging time-scale varies linearly from 3 days at the boundary to 10 days at the interior end of the 15 grid point wide nudging zone in all 24 vertical layers. A sponge layer with enhanced diffusion of tracers and momentum in the northernmost 10 grid points minimizes reflections at the northern channel wall, and a full-depth nudging of temperature and salinity (with a 30 day time scale) in the sponge layer is applied to preserve a horizontally homogeneous water mass distribution in the deep ocean. The surface properties Diflunisal outside the FIS are largely determined by the annul cycle of melting and freezing of sea ice

(Nicholls et al., 2009). To mimic the effect of sea ice, which is not included in our model, temperature and salinity within the uppermost model layer are directly restored to the horizontally averaged surface climatology obtained from the seal data, with a nudging time scale of 10 days. This setup for the hydrographic forcing avoids the uncertainties associated with poorly constrained fluxes at the air-ice-ocean boundary, and allows us to study the direct oceanic response to different upper ocean conditions, while assuring a consistent model forcing. For the mechanical surface forcing, a wind stress that is constant in time, but resolves the average spatial pattern of the wind field in the model domain is applied. The forcing field is derived by time-averaging the RACMO2 results, with minor modifications applied in order to ensure periodicity at the boundaries.