Berlin 2012 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
SOE: Fachverband Physik sozio-ökonomischer Systeme
SOE 9: Energy and Environment
SOE 9.4: Vortrag
Dienstag, 27. März 2012, 12:00–12:15, H 0110
On the Predictability of El Niño by Climate Networks — •Josef Ludescher1, Avi Gozolchiani2, Mikhail Bogachev1, Shlomo Havlin2, and Armin Bunde1 — 1Institut für Theoretische Physik III, Justus Liebig Universität Gießen, 35392 Gießen, Germany — 2Minerva Center and Department of Physics, Bal Ilan University, Ramat Gan, Israel
We construct and analyze climate networks based on gridded observational data starting 1948. The grid sites form the nodes of the network and links represent cooperative behavior between these nodes. We define the link weight as the strength of the crosscorrelation between the time series at the nodes, which can be temperature, pressure, wind data etc. We find that the surface air temperature links that connect the El Niño basin with the rest of the equatorial Pacific are the most sensitive for El Niño and show that the network strength of this area, defined as the sum of the link weights, in most cases precedes the El Niño Index by several months. We use ROC analysis to show that most El Niño events and non-events can be predicted in the preceding year.