 |
| Titel |
Lyapunov covariant modes and predictability |
| VerfasserIn |
Bernard Legras, Guillaume Lapeyre, Pietro Peterlongo |
| Konferenz |
EGU General Assembly 2011
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 13 (2011) |
| Datensatznummer |
250049967
|
|
|
|
|
|
| Zusammenfassung |
| The dynamics of the atmosphere at mid-latitudes is characterized by its
varying predictability, that is to say that there exists atmospheric
regimes that are more predictable (for which errors in initial
conditions grow less rapidly) than others. The error growth can be
measured using by singular values and singular vectors for a finite
time integration, or using Lyapunov exponents and Lyapunov vectors
that generalize this notion for time-dependent dynamical systems in
the asymptotic limit.
The spatial heterogeneity of the Lyapunov vectors is associated with
localization properties of the instabilities and is related to the variations
in space or time of the predictability. However this notion is difficult to apprehend
using singular vectors because they rely on the choice of the norm.
On the contrary, the covariant vectors which are preserved by the flow
and are independent of any norm, bears desirable properties for
predictability studies. We review recent developments in the
computation of these covariant vectors and give some views on their
properties in simple dynamical and turbulent models. |
|
|
|
|
|
|