HR: 11:10h
AN: NG32A-04 INVITED    [Abstracts]
TI: Systematic Identification of Metastable Regimes in Atmospheric Data Sets
AU: * Franzke, C
EM: chan1@bas.ac.uk
AF: British Antarctic Survey, High Cross, Madingley Road, Cambridge, CB3 0ET, United Kingdom
AU: Horenko, I
EM: horenko@math.fu-berlin.de
AF: Free University Berlin, Arnimallee 2-6, Berlin, 14195, Germany
AU: Majda, A J
EM: majda@cims.nyu.edu
AF: Courant Institute of Mathematical Sciences, New York University, 251 Mercer Street, New York, NY 10012, United States
AB: A pronounced characteristic of the atmospheric circulation is its irregularity, visible in the daily change of the weather. Despite this chaotic behaviour it is well known that certain flow structures tend to occur over and over again. These recurring flow structures are commonly called atmospheric flow regimes and inspired a whole body of work. In this talk I will present a novel method which simultanously estimates possible metastable states, the Markov transition matrix for the switching between the metastable states and the corresponding local Principal Components. This methodology is based on combining Hidden Markov Models and Principal Component Analysis. If the Markov transition matrix possesses metastable (or quasi-persistent) states, we identify these as regimes. In this perspective, regimes can be present even though the observed data has a nearly Gaussian (unimodal) probability distribution. We apply this procedure to data from a model of barotropic flow over topography with a large scale mean flow. This model exhibits regime behaviour for sufficiently high topography. The regime structure and their dynamical significance will be discussed.
DE: 3265 Stochastic processes (3235, 4468, 4475, 7857)
SC: Nonlinear Geophysics [NG]
MN: 2007 Fall Meeting