390
irradiance changes have an amplitude of less than 2 Wm- 2 (ERBE 1990)
which would yield an amplitude of about 0.3 K using the above sensitivity.
It is therefore important to look for possible alternative explanations and
mechanisms of climate variability.
Modeling has become an important branch of climate research because
only with physically based models is it possible to quantitatively test and
verify hypotheses on climate change. Ocean, atmosphere and coupled models are successfully simulating the large-scale fields of the climate system
(Trenberth 1992), and it is now timely to assess these models' capability
of simulating also the natural variability.
During the last decade, oceanic circulation models have made significant
progress due to the implementation of mixed boundary conditions that take
into account feedback mechanisms between atmosphere and ocean. SST
anomalies generate local heat flux anomalies that operate to remove the
SST anomalies within a few weeks. Sea surface salinity (SSS) anomalies,
on the other hand, do not influence the hydrological cycle, i.e. the surface
freshwater balance, and hence can have a longer lasting impact on the surface buoyancy distribution. By relaxing SST to a fixed surface air temperature and keeping the surface freshwater fluxes constant, one arrives at a first
approximation of the important difference between the feedback character
of ocean-to-atmosphere heat and freshwater fluxes (StommeI1961; Rooth
1982). However, local heat flux anomalies are bound to also change surface
air temperature, an effect which is explicitly excluded when using mixed
boundary conditions. Recognizing these limitations several studies have
proposed improved parameterizations of the surface fluxes by formulating
various types of energy balance models coupled to the ocean circulation
models (e.g. Stocker et al. (1992), Zhang et ai. (1993), Rahmstorf and
Willebrand (1995), Lohmann et al. (1996)). Two effects are of importance:
(i) SST anomalies cause heat flux anomalies which modify the surface air
temperature locally; (ii) due to the possibility of meridional heat flux in
the atmospheric part of the coupled model far-field effects can occur.
The basic mechanism for oscillations due to different feedback processes of
SST and SSS anomalies was summarized by Welander (1986). He showed
that self-sustained oscillations and different equilibrium states can be realized in a circular convection loop in which one side is heated and salted
while the other side is cooled and freshened. This is reminiscent of the
low and high latitudes where the surface ocean is heated and evaporation
causes. an increase in salinity whereas the opposite happens in the high
irradiance changes have an amplitude of less than 2 Wm- 2 (ERBE 1990)
which would yield an amplitude of about 0.3 K using the above sensitivity.
It is therefore important to look for possible alternative explanations and
mechanisms of climate variability.
Modeling has become an important branch of climate research because
only with physically based models is it possible to quantitatively test and
verify hypotheses on climate change. Ocean, atmosphere and coupled models are successfully simulating the large-scale fields of the climate system
(Trenberth 1992), and it is now timely to assess these models' capability
of simulating also the natural variability.
During the last decade, oceanic circulation models have made significant
progress due to the implementation of mixed boundary conditions that take
into account feedback mechanisms between atmosphere and ocean. SST
anomalies generate local heat flux anomalies that operate to remove the
SST anomalies within a few weeks. Sea surface salinity (SSS) anomalies,
on the other hand, do not influence the hydrological cycle, i.e. the surface
freshwater balance, and hence can have a longer lasting impact on the surface buoyancy distribution. By relaxing SST to a fixed surface air temperature and keeping the surface freshwater fluxes constant, one arrives at a first
approximation of the important difference between the feedback character
of ocean-to-atmosphere heat and freshwater fluxes (StommeI1961; Rooth
1982). However, local heat flux anomalies are bound to also change surface
air temperature, an effect which is explicitly excluded when using mixed
boundary conditions. Recognizing these limitations several studies have
proposed improved parameterizations of the surface fluxes by formulating
various types of energy balance models coupled to the ocean circulation
models (e.g. Stocker et al. (1992), Zhang et ai. (1993), Rahmstorf and
Willebrand (1995), Lohmann et al. (1996)). Two effects are of importance:
(i) SST anomalies cause heat flux anomalies which modify the surface air
temperature locally; (ii) due to the possibility of meridional heat flux in
the atmospheric part of the coupled model far-field effects can occur.
The basic mechanism for oscillations due to different feedback processes of
SST and SSS anomalies was summarized by Welander (1986). He showed
that self-sustained oscillations and different equilibrium states can be realized in a circular convection loop in which one side is heated and salted
while the other side is cooled and freshened. This is reminiscent of the
low and high latitudes where the surface ocean is heated and evaporation
causes. an increase in salinity whereas the opposite happens in the high