 |
| Titel |
Do climate simulations from models forced by averaged sea surface temperatures represent actual dynamics? |
| VerfasserIn |
P. J. Roebber, A. A. Tsonis  , J. B. Elsner |
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| ISSN |
1023-5809
|
| Digitales Dokument |
URL |
| Erschienen |
In: Nonlinear Processes in Geophysics ; 4, no. 2 ; Nr. 4, no. 2, S.93-100 |
| Datensatznummer |
250001575
|
| Publikation (Nr.) |
 copernicus.org/npg-4-93-1997.pdf |
|
|
|
|
|
| Zusammenfassung |
| Recently atmospheric general circulation models
(AGCMs)
forced by observed sea surface temperatures (SSTs) have offered the possibility
of studying climate variability over periods ranging from years to decades.
Such models represent and alternative to fully coupled asynchronous atmosphere
ocean models whose long term integration remains problematic. Here, the
degree of the approximation represented by this approach is investigated
from a conceptual point of view by comparing the dynamical properties of
a low order coupled atmosphere-ocean model to those of the atmospheric
component of the same model when forced with monthly values of SST derived
from the fully coupled simulation. The low order modelling approach is undertaken
with the expectation that it may reveal general principles concerning the
dynamical behaviour of the forced versus coupled systems; it is not expected
that such an approach will determine the details of these differences,
for which higher order modelling studies will be required. We discover that
even though attractor (global) averages may be similar, local dynamics
and the resultant variability and predictability characteristics differ
substantially. These results suggest that conclusions concerning regional
climatic variability (in time as well as space) drawn from forced modelling
approaches may be contaminated by an inherently unquantifiable error. It
is therefore recommended that this possibility be carefully investigated
using state-of-the-art coupled AGCMs. |
| |
|
| Teil von |
|
|
|
|
|
|
|
|