Dresden 2009 – wissenschaftliches Programm

Bereiche | Tage | Auswahl | Suche | Downloads | Hilfe

DY: Fachverband Dynamik und Statistische Physik

DY 27: Poster II

DY 27.7: Poster

Donnerstag, 26. März 2009, 16:00–18:00, P1B

Local Low Dimensionality and Predictability in the Lorenz96 Model of Atmospheric Dynamics — •Stefan Siegert and Günter Radons — Chemnitz University of Technology, Chemnitz, Germany

The Lorenz96 model is a prototype model mimicking certain features of mid-latitude atmospheric dynamics. It is widely used for conceptual studies of atmospheric predictability. Bred vectors characterize finite-time dynamics of high-dimensional chaotic systems under the influence of finite-size perturbations. In the present study, Bred vectors of the Lorenz96 model are characterized in terms of localization, convergence and propagation and compared to their infinitesimal-size, infinite-time counterpart, the leading Lyapunov vectors.

In recent studies it was found, that Bred vectors in atmospheric models exhibit anomalously low dimensionality in localized regions. This finding might have important implications for predictability. In the present study, regions of low Bred vector dimensionality in the Lorenz96 model were identified. There is evidence, that short-term predictability, defined in terms of information entropy of an ensemble of solutions, is low (i. e. entropy is high) in regions of low Bred vector dimension. Presently, further connections between local low dimensionality and predictability are established.