pace of the two stages in the adjustment process
that provides the main rationale for model integration times of a few decades, and motivation for the
analyses of model solutions in dynamic quasiequilibrium in relation to observed ocean behaviours.
While simulations of thermodynamic equilibrium
are faced with the difficulty of an inaccurate
knowledge of the history of surface fluxes over the
secular time scales of required integration periods,
simulations of the dynamical quasi-equilibrium
stage can increasingly capitalize on improvements
in both initial conditions (i.e. through the synthesis
of the WOCE hydrographic database) and atmospheric fluxes (e.g. through the re-analysis projects
in the major meteorological forecasting centres).
Accordingly, a quantitative comparison with
observed time series of dynamical aspects of ocean
circulation, such as surface height and boundary
current transports, is rapidly gaining importance
as a most powerful tool for an evaluation of model
performance. It should be noted, however, that all
such model analyses and confrontation of solutions with data have to account for the fact that
the hydrographic (property) values in the deep
ocean interior, away from intense currents, are to
a large degree reflecting the initial conditions: for
those areas a direct comparison with subthermocline WOCE section data is of little use. Nevertheless, the thermocline properties of WOCE sections
and XBT (Expendable bathythermograph) lines
allow valuable comparisons with the simulated
vertical structures in models; and deep drifters
provide intriguing zonal flows against which to
evaluate model velocities and investigate dynamical mechanisms. In any event, the drift in the deep
water mass properties, albeit weak, is not zero,
implying a continuing, although very slow, evolution of the associated circulation and property
fluxes. (For example, due to the ongoing storing of
heat after the initial, dynamic adjustment phase, it
requires integration times of several more decades
to reduce local imbalances between the surface
flux and the oceanic transport of heat below
0.1 PW.) The presence of the slow drift means
that model property fluxes after a few decades
of integration have to be interpreted with due
caution.
2.2.4 Examples of model behaviour in
different dynamical regimes
2.2.4.1 Wind-driven circulation and its
variability on intraseasonal to interannual
time scales
Model runs forced with a history of realistic timevarying atmospheric fluxes have shown increasing
success in recent years in quantitatively reproducing time series from altimetric data, tide gauges,
and in-situ station data such as western boundary
current transport variability. The major forcing
field for ocean current variability on intraseasonal,
seasonal, and, to some extent, interannual time
scales is the wind stress. Earlier ocean models typically relied on seasonal climatologies of surface
fluxes estimated from historical marine observations and bulk aerodynamic formulae. In recent
SECTION 2 OBSERVATIONS AND MODELS
64
Transport (Sv)
Heat transport (PW)
t –t n (years)
t – t 0 (years)
(a)
(b)
Fig. 2.2.2 Response of large-scale transport in the
North Atlantic to a sudden change in the prescribed
hydrographic properties of the Denmark Strait Overflow
Water, from the model study of Döscher et al. (1994).
(a) Zonally integrated transport of the NADWoverturning cell and (b) of poleward heat transport, at
selected latitudes.Two model cases (C2, C3) are shown,
both initialized (t:0) with the equilibrated state of a
reference case. In C2 a denser outflow was imposed,
while in C3 no outflow effect was included.
that provides the main rationale for model integration times of a few decades, and motivation for the
analyses of model solutions in dynamic quasiequilibrium in relation to observed ocean behaviours.
While simulations of thermodynamic equilibrium
are faced with the difficulty of an inaccurate
knowledge of the history of surface fluxes over the
secular time scales of required integration periods,
simulations of the dynamical quasi-equilibrium
stage can increasingly capitalize on improvements
in both initial conditions (i.e. through the synthesis
of the WOCE hydrographic database) and atmospheric fluxes (e.g. through the re-analysis projects
in the major meteorological forecasting centres).
Accordingly, a quantitative comparison with
observed time series of dynamical aspects of ocean
circulation, such as surface height and boundary
current transports, is rapidly gaining importance
as a most powerful tool for an evaluation of model
performance. It should be noted, however, that all
such model analyses and confrontation of solutions with data have to account for the fact that
the hydrographic (property) values in the deep
ocean interior, away from intense currents, are to
a large degree reflecting the initial conditions: for
those areas a direct comparison with subthermocline WOCE section data is of little use. Nevertheless, the thermocline properties of WOCE sections
and XBT (Expendable bathythermograph) lines
allow valuable comparisons with the simulated
vertical structures in models; and deep drifters
provide intriguing zonal flows against which to
evaluate model velocities and investigate dynamical mechanisms. In any event, the drift in the deep
water mass properties, albeit weak, is not zero,
implying a continuing, although very slow, evolution of the associated circulation and property
fluxes. (For example, due to the ongoing storing of
heat after the initial, dynamic adjustment phase, it
requires integration times of several more decades
to reduce local imbalances between the surface
flux and the oceanic transport of heat below
0.1 PW.) The presence of the slow drift means
that model property fluxes after a few decades
of integration have to be interpreted with due
caution.
2.2.4 Examples of model behaviour in
different dynamical regimes
2.2.4.1 Wind-driven circulation and its
variability on intraseasonal to interannual
time scales
Model runs forced with a history of realistic timevarying atmospheric fluxes have shown increasing
success in recent years in quantitatively reproducing time series from altimetric data, tide gauges,
and in-situ station data such as western boundary
current transport variability. The major forcing
field for ocean current variability on intraseasonal,
seasonal, and, to some extent, interannual time
scales is the wind stress. Earlier ocean models typically relied on seasonal climatologies of surface
fluxes estimated from historical marine observations and bulk aerodynamic formulae. In recent
SECTION 2 OBSERVATIONS AND MODELS
64
Transport (Sv)
Heat transport (PW)
t –t n (years)
t – t 0 (years)
(a)
(b)
Fig. 2.2.2 Response of large-scale transport in the
North Atlantic to a sudden change in the prescribed
hydrographic properties of the Denmark Strait Overflow
Water, from the model study of Döscher et al. (1994).
(a) Zonally integrated transport of the NADWoverturning cell and (b) of poleward heat transport, at
selected latitudes.Two model cases (C2, C3) are shown,
both initialized (t:0) with the equilibrated state of a
reference case. In C2 a denser outflow was imposed,
while in C3 no outflow effect was included.