Different processes have been proposed to explain the large-scale spreading
of Mediterranean Water (MW) in the North Atlantic, however, no systematic
study comparing the efficiency of different processes is yet available. Her
e, the authors present a series of experiments in a unified framework that
is designed to quantify the effects of several physical processes on the sp
reading of MW in an idealized model of the North Atlantic. The common techn
ique of restoring temperature and salinity to an observed distribution near
the Mediterranean inflow Fails to produce an adequate amount of MW because
the eastern boundary region near the MW inflow is rather quiescent in mode
ls. Diapycnal processes like double diffusion and cabbeling turn out too in
efficient to alone account for the large-scale MW anomaly. However, with a
preexisting anomaly, double diffusion leads to a considerable northward and
zonal redistribution of MW. The density anomaly induced by cabbeling curta
ils the zonal spreading of MW while it increases the northward spreading, W
ith isopycnal mixing and the weak mean flow that prevails in the outflow re
gion, a spatial distribution of the MW anomaly is obtained that is inconsis
tent with observations. Unrealistically high diffusion coefficients would b
e necessary to reproduce the observed salt flux into the Atlantic. The most
effective process in the experiments is the volume Flux associated with th
e Atlantic-Mediterranean exchange. The current system that is established i
n response to the inflow of MW into the Atlantic carries the anomaly almost
30 degrees of longitude into the basin and along the eastern margin up to
the northeastern corner of the domain and farther along the northern bounda
ry.