568 JOURNAL OF THE ATMOSPHERIC SCIENCES VOLUME 61 Weather Regimes and Preferred Transition Paths in a Three-Level Quasigeostrophic Model

Multiple flow regimes are reexamined in a global, three-level, quasigeostrophic (QG3) model with realistic topography in spherical geometry. This QG3 model, using a T21 triangular truncation in the horizontal, has a fairly realistic climatology for Northern Hemisphere winter and exhibits multiple re...

Full description

Bibliographic Details
Main Authors: D. Kondrashov, K. Ide, M. Ghil
Other Authors: The Pennsylvania State University CiteSeerX Archives
Format: Text
Language:English
Published: 2003
Subjects:
Online Access:http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.332.7372
http://www.atmos.ucla.edu/~kayo/data/publication/kondrashov_etal_jas04.pdf
Description
Summary:Multiple flow regimes are reexamined in a global, three-level, quasigeostrophic (QG3) model with realistic topography in spherical geometry. This QG3 model, using a T21 triangular truncation in the horizontal, has a fairly realistic climatology for Northern Hemisphere winter and exhibits multiple regimes that resemble those found in atmospheric observations. Four regimes are robust to changes in the classification method, k-means versus mixture modeling, and its parameters. These regimes correspond roughly to opposite phases of the Arctic Oscillation (AO) and the North Atlantic Oscillation (NAO), respectively. The Markov chain representation of regime transitions is refined here by finding the preferred transition paths in a three-dimensional (3D) subspace of the model’s phase space. Preferred transitions occur from the positive phase of the NAO (NAO � ) to that of the AO (AO �), from AO � to NAO � , and from NAO � to NAO � , but not directly between opposite phases of the AO. The angular probability density function (PDF) of the regime exits that correspond to these preferred transitions have one or, sometimes, two fairly sharp maxima. These angular PDF maxima are, in most cases, not aligned with the line segments between regime centroids in phase space and might point to heteroclinic or homoclinic connections between unstable equilibria in the model’s phase space. Preferred transitions paths are also determined for a stochastically forced Lorenz system to help explain