402
are essentially red noise, and there appears an out-of-phase relationship
between the stratosphere and the troposphere. In the ocean there is an
irregular cycle of about 17 years located in the Pacific Ocean. Note that
no distinct variability is found in the Atlantic basin neither in marginal
seas nor basins-wide.
4 Conclusions
The study and quantitative understanding of internal variability in the
climate system is still in its infancy. Focal regions where internal variability
is enhanced or generated have been identified and preliminary results have
shed light on a palette of mechanisms. However, we are still at the stage
where each model tends to produce its characteristic set of fluctuations with
the associated time scales. Important results have been found regarding
the role of the ocean, the influence of atmosphere-ocean exchange fluxes
and the dynamics in marginal seas, but a consistent quantitative theory on
climatic cycles on the decadal-to-century time scale is still missing. Further
work is urgently needed both with respect of models and observations. We
need to improve the formulation of surface exchange processes (boundary
conditions) especially in the ocean-only models, the parameterisation of
convection and deep water formation; the resolution must be increased in
order to better represent topography and marginal but important regions
of the ocean basins must be included.
More important, however, are efforts to obtain and analyze high-quality
terrestrial and oceanic proxy data which allow annual or seasonal resolution. Special attention must be focused on the transfer function, i.e. how
is the climate signal transferred into the archive and how is the possible
temporal evolution of such a transfer function. Ongoing European and
US efforts to retrieve two high-resolution ice cores from different locations
in Antarctica will provide vital information about natural variability centered in the Southern Ocean. Finally, spatial networks of homogenized
paleoclimatic data will certainly allow us to find some of the keys to better
understand natural variability in the climate system.
Acknowledgment: I thank Jiirgen Willebrand and David Anderson for this
stimulating NATO ASI workshop held at Les Houches in February 1995.
Thomas Blunier compiled Figure 2. I thank Thomas Tschannen for his
expert help with rnFjX and an anonymous reviewer for comments which
clarified the distinction between different types of climate variability. This
work is supported by the Swiss National Science Foundation.
are essentially red noise, and there appears an out-of-phase relationship
between the stratosphere and the troposphere. In the ocean there is an
irregular cycle of about 17 years located in the Pacific Ocean. Note that
no distinct variability is found in the Atlantic basin neither in marginal
seas nor basins-wide.
4 Conclusions
The study and quantitative understanding of internal variability in the
climate system is still in its infancy. Focal regions where internal variability
is enhanced or generated have been identified and preliminary results have
shed light on a palette of mechanisms. However, we are still at the stage
where each model tends to produce its characteristic set of fluctuations with
the associated time scales. Important results have been found regarding
the role of the ocean, the influence of atmosphere-ocean exchange fluxes
and the dynamics in marginal seas, but a consistent quantitative theory on
climatic cycles on the decadal-to-century time scale is still missing. Further
work is urgently needed both with respect of models and observations. We
need to improve the formulation of surface exchange processes (boundary
conditions) especially in the ocean-only models, the parameterisation of
convection and deep water formation; the resolution must be increased in
order to better represent topography and marginal but important regions
of the ocean basins must be included.
More important, however, are efforts to obtain and analyze high-quality
terrestrial and oceanic proxy data which allow annual or seasonal resolution. Special attention must be focused on the transfer function, i.e. how
is the climate signal transferred into the archive and how is the possible
temporal evolution of such a transfer function. Ongoing European and
US efforts to retrieve two high-resolution ice cores from different locations
in Antarctica will provide vital information about natural variability centered in the Southern Ocean. Finally, spatial networks of homogenized
paleoclimatic data will certainly allow us to find some of the keys to better
understand natural variability in the climate system.
Acknowledgment: I thank Jiirgen Willebrand and David Anderson for this
stimulating NATO ASI workshop held at Les Houches in February 1995.
Thomas Blunier compiled Figure 2. I thank Thomas Tschannen for his
expert help with rnFjX and an anonymous reviewer for comments which
clarified the distinction between different types of climate variability. This
work is supported by the Swiss National Science Foundation.