85
ity, and indicate that this effect is much less predominant in the tropics.
In doing this a new interpretation of the Charney-Shukla paradigm for
large-scale tropical predictability is given.
In section 4, we discuss the predictability of forced problems (predictability of the second kind). Explicit examples are given using both the
Lorenz 3-component model, and the global circulation models (GCMs).
We develop a basic nonlinear paradigm for analysing the response to an
external perturbation. The basic notion is that the influence of a weak
external forcing is greatest in regions of phase space where the system is
particularly unstable; however, the response of the system is greatest in
regions of phase space where the system is particlarly stable. The extratropical response of the atmosphere to extratropical sea surface temperature
(SST) anomalies is discussed using this paradigm.
In section 5, the predictability of natural fluctuations of the climate
system on decadal timescales is discussed on the basis of a number of GCM
integrations. One fundamental question concerns the relative contribution
of the atmospheric and oceanic dyamics in contributing to observed decadal
variability. It is found that, even on decadal timescales, the role of purely
internal atmospheric variability is not negligible. This itself implies that
the predictability of decadal fluctuations may not be strong. However, in
addition, it is suggested that the role of the extratropical oceans is basically
to redden the spectrum of atmospheric variability, eg by increasing the
typical residence time of the atmospheric state vector within a weather
regime. As such, decadal atmospheric variability linked directly to decadal
SST variability may not itself be strongly predictable.
In section 6, the nonlinear paradigm discussed earlier is applied to the
problem of climate change. We assess whether or not the observed warming of the atmosphere can be attributed to the greenhouse effect. It is
suggested that the observed warming over the past few decades can be
largely interpreted in terms of an increase in the frequency of one of the
dominant regimes of the extratropical flow. Singular vector analysis suggests that this regime may be most sensitive to forcing in the tropical west
Pacific warm pool region. This therefore may be the most critical area
for understanding how enhanced CO2 may influence global climate. The
predictability implications of this analysis are discussed.
In section 7, some remarks are made about the rationalisation of climate
and weather prediction models.
In this paper, we shall make some use of low-dimensional chaotic models.
ity, and indicate that this effect is much less predominant in the tropics.
In doing this a new interpretation of the Charney-Shukla paradigm for
large-scale tropical predictability is given.
In section 4, we discuss the predictability of forced problems (predictability of the second kind). Explicit examples are given using both the
Lorenz 3-component model, and the global circulation models (GCMs).
We develop a basic nonlinear paradigm for analysing the response to an
external perturbation. The basic notion is that the influence of a weak
external forcing is greatest in regions of phase space where the system is
particularly unstable; however, the response of the system is greatest in
regions of phase space where the system is particlarly stable. The extratropical response of the atmosphere to extratropical sea surface temperature
(SST) anomalies is discussed using this paradigm.
In section 5, the predictability of natural fluctuations of the climate
system on decadal timescales is discussed on the basis of a number of GCM
integrations. One fundamental question concerns the relative contribution
of the atmospheric and oceanic dyamics in contributing to observed decadal
variability. It is found that, even on decadal timescales, the role of purely
internal atmospheric variability is not negligible. This itself implies that
the predictability of decadal fluctuations may not be strong. However, in
addition, it is suggested that the role of the extratropical oceans is basically
to redden the spectrum of atmospheric variability, eg by increasing the
typical residence time of the atmospheric state vector within a weather
regime. As such, decadal atmospheric variability linked directly to decadal
SST variability may not itself be strongly predictable.
In section 6, the nonlinear paradigm discussed earlier is applied to the
problem of climate change. We assess whether or not the observed warming of the atmosphere can be attributed to the greenhouse effect. It is
suggested that the observed warming over the past few decades can be
largely interpreted in terms of an increase in the frequency of one of the
dominant regimes of the extratropical flow. Singular vector analysis suggests that this regime may be most sensitive to forcing in the tropical west
Pacific warm pool region. This therefore may be the most critical area
for understanding how enhanced CO2 may influence global climate. The
predictability implications of this analysis are discussed.
In section 7, some remarks are made about the rationalisation of climate
and weather prediction models.
In this paper, we shall make some use of low-dimensional chaotic models.