 |
| Titel |
Atmospheric properties and the ENSO cycle |
| VerfasserIn |
S. Y. Philip, G. J. van Oldenborgh |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250021477
|
|
|
|
|
|
| Zusammenfassung |
| This study systematically characterizes the main feedbacks in the ENSO cycle and
(nonlinear) atmospheric properties with the help of a reduced model tuned to the couplings of
reanalyses or GCM output. Most climate models are found to underestimate the response of
SST to wind stress anomalies, which is compensated by a damping term that is
lower than observed. Furthermore they do not show the correct SST skewness and
underestimate the wind variability and the nonlinear response of wind stress to
SST.
First, the linear dynamics of ENSO is examined from observations. These are the
relations between SST, zonal wind stress (τx) and the thermocline depth or 20∘C isotherm in
the ocean (Z20). In addition some nonlinear terms in the atmosphere are studied. These are
the nonlinear response of τx to SST, the standard deviation and skewness of the component
of the noise in τx (the component that is not part of the ENSO cycle) and the relation between
background SST and this wind noise.
An Intermediate Complexity Model (ICM) is built with these findings. This reduced
model in which all couplings are fitted to observations can simulate the main properties of
ENSO. In this initially linear model we systematically introduce the extra terms in the
atmospheric component. We study the influence of these terms on ENSO properties like
amplitude, period and skewness from this ICM.
Then we investigate the differences between observations and coupled general circulation
models (CGCMs). We compare the linear dynamics of ENSO in CGCMs with observations
in the models with the best linear dynamics. For these models we compare the
nonlinear terms with observations. The influence of differences between the CGCMs
and observations on ENSO properties is examined with the ICM, now fitted to
the CGCMs. The reduced model can simulate the main properties of ENSO in
GCMs.
Our reduced model can be used to trace back those changes to physical coupling
parameters, and further back to model parameters. This will allow models to be tuned for
more realistic ENSO simulations, enhancing the skill of seasonal forecasts and increasing our
confidence in climate change projections. |
|
|
|
|
|
|