They found that prescribing enhanced vertical diffusion slows

the downslope progression of the plume, while prescribing enhanced vertical viscosity increases downslope transport (given sufficient supply of dense water). The agreement with the descent rate prediction of Shapiro and Hill (1997) was shown by Wobus et al. (2011) not to be limited to cascades with a sharp interface and a thin plume with hF∼O(He)hF∼O(He), but also applicable to thick and diffuse plumes as long as the vertical diffusivity κκ and viscosity PLX3397 solubility dmso ν   are of approximately the same magnitude (i.e. a vertical Prandtl number of Prv∼O(1)Prv∼O(1)). This study confirms the ( Shapiro and Hill (1997)) descent rate formula in a model using the GLS turbulence closure scheme (rather than prescribed turbulence). The agreement in Fig. 7 is explained by plumes of the ‘piercing’ regime of our experiments meeting the aforementioned Prandtl number criterion (see Table 1). On its downslope descent the plume (SFOW) mixes with and entrains three ambient water masses (ESW, AW and NSDW). Entrainment implying a volume click here increase is based on a potentially arbitrary distinction between plume water and ambient water which could result in imprecise heat and salt budgets. In the following we therefore concentrate on the mixing process where these budgets remain

well defined. Fig. 8 shows θ-S diagrams that trace the water properties down the slope at the end of each experiment (after 90 days). The θ-S values are plotted for the bottom model level at increasing depths from inflow region down to 1500 m. We show the θ-S properties for two experiments series: Q is ID-8 constant and S varies ( Fig. 8(a)), and Q varies and S is constant ( Fig. 8(b)). The dashed portion of the mixing curves in Fig. 8 shows that a considerable amount of mixing takes place within the injection grid cells. Any water introduced into the model is immediately diluted by

ambient water. These processes take place over a very small region of the model and are not considered any further. Instead we focus on the common feature of all curves in Fig. 8: the temperature rises to a temperature maximum (marked by red squares) due to the plume’s mixing with warm Atlantic Water. A very similar mixing characteristic was described by Fer and Ådlandsvik (2008) for a single overflow scenario ( S=35.3,T=-1.9°C,Qavg=0.07Sv) in a 3-D model study using ambient conditions similar to ours. Amongst the series with constant Q  =0.03 Sv ( Fig. 8(a)) only the weakest cascade (inflow salinity S  =34.75) retains traces of ESW in the bottom layer after 90 days. In the experiments with more saline inflow (S⩾35.00S⩾35.00), the θ-S curve in Fig. 8(a) only spans three water masses – SFOW, AW and NSDW – while ESW is no longer present near the seabed. The salinity at the temperature maximum is nearly identical (red squares in Fig. 8(a)) for runs with the same flow rate Q. The experiments with a constant inflow salinity S ( Fig.