Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

DY: Fachverband Dynamik und Statistische Physik

DY 26: Nonlinear Dynamics II

DY 26.6: Vortrag

Mittwoch, 16. März 2011, 17:45–18:00, ZEU 255

Conditional Mutual Sorting Information for Coupling Analysis of Time Series — •Jakob Runge¹ and Bernd Pompe² — ¹Potsdam-Institut für Klimafolgenforschung — ²Universität Greifswald

We propose a method to detect couplings between two simultaneously measured time series. It is based on conditional mutual sorting information. By setting suitable conditions, we are able to diminish misleading effects of auto dependencies within each series. This enables the detection of the right coupling delays also in the case of bidirectional couplings. Our approach is based on ordinal properties of time series. This makes the analysis invariant with respect to monotonous distortions which is very useful, e.g., in the analysis of proxy data in climatology. Moreover, ordinal analysis is robust to some trend, and easy and fast to compute. We consider also the problem of reliable estimation from finite time series. Finally, we apply the proposed method to nonlinear models as well as to some climate data.