149
time, and therefore only cover specific synoptic situations. A comparison
of COARE fluxes with some mean climatology from a long integration of
a climate model is clearly inadequate. In comparing model with data, the
model should be run in forecast mode. Short-range forecast errors in fluxes
can then be compared directly with the observations, and the impact of
revisions to the model physics (eg to cloud parametrisation schemes) can be
readily validated. By validating the model for predictions of the first kind,
we should have improved the model's reliability in making predictions of
the second kind.
In addition to these concerns, we need to be able to pool modelling
resources in order to test hypotheses in a fashion in which problems of
model dependence can be minimised. The Atmospheric Model Intercomparison Project (Gates, 1993) is a good example of collaborative research
that highlights the successes and limitations of the current generation of
models.
However, it has been more difficult, on the basis of these intercomparison
studies, to provide corrective prescriptions that will reduce the overall level
of model bias. In my opinion, a fundamental question that the climate
community needs to address, is whether we can achieve the required level
of model accuracy for reliable climate prediction through the development
of a diverse range of GCMs, maintained on an individual institute basis,
or whether resources should be pooled into the detailed investigation and
development of a smaller subset of proven models. I have my views (Palmer
and Webster, 1995), I would like to hear yours!
Acknow ledgernents
I would like to thank Jan Barkmeijer, David Battisti, Roberto Buizza,
Ying-Quei Chen, Ron Gelaro, David Rowell and Stephen Griffies for providing me with material which has yet to be published. I am also grateful
to Andy Moore and David Stephenson for very helpful comments on an
earlier version of the manuscript.
References
ANGELL, J.K., 1988, Variations and trends in tropospheric and stratospheric global temperatures. 1958-87. J.Clim., 1, 1296-1313.
ABARBANEL, H.D.I., R.BROWN, AND M.B.KENNEL, 1991, 'Variation of Lyapunov exponents on a strange attractor', Journal of Nonlinear Science, 1, 175-199.
BALMADEDA, M.A., M.K.DAVEY, AND D.L.T.ANDERSON, 1995, Decadal and seasonal dependence of ENSO prediction skill. J.Climate.; In press.
time, and therefore only cover specific synoptic situations. A comparison
of COARE fluxes with some mean climatology from a long integration of
a climate model is clearly inadequate. In comparing model with data, the
model should be run in forecast mode. Short-range forecast errors in fluxes
can then be compared directly with the observations, and the impact of
revisions to the model physics (eg to cloud parametrisation schemes) can be
readily validated. By validating the model for predictions of the first kind,
we should have improved the model's reliability in making predictions of
the second kind.
In addition to these concerns, we need to be able to pool modelling
resources in order to test hypotheses in a fashion in which problems of
model dependence can be minimised. The Atmospheric Model Intercomparison Project (Gates, 1993) is a good example of collaborative research
that highlights the successes and limitations of the current generation of
models.
However, it has been more difficult, on the basis of these intercomparison
studies, to provide corrective prescriptions that will reduce the overall level
of model bias. In my opinion, a fundamental question that the climate
community needs to address, is whether we can achieve the required level
of model accuracy for reliable climate prediction through the development
of a diverse range of GCMs, maintained on an individual institute basis,
or whether resources should be pooled into the detailed investigation and
development of a smaller subset of proven models. I have my views (Palmer
and Webster, 1995), I would like to hear yours!
Acknow ledgernents
I would like to thank Jan Barkmeijer, David Battisti, Roberto Buizza,
Ying-Quei Chen, Ron Gelaro, David Rowell and Stephen Griffies for providing me with material which has yet to be published. I am also grateful
to Andy Moore and David Stephenson for very helpful comments on an
earlier version of the manuscript.
References
ANGELL, J.K., 1988, Variations and trends in tropospheric and stratospheric global temperatures. 1958-87. J.Clim., 1, 1296-1313.
ABARBANEL, H.D.I., R.BROWN, AND M.B.KENNEL, 1991, 'Variation of Lyapunov exponents on a strange attractor', Journal of Nonlinear Science, 1, 175-199.
BALMADEDA, M.A., M.K.DAVEY, AND D.L.T.ANDERSON, 1995, Decadal and seasonal dependence of ENSO prediction skill. J.Climate.; In press.