Predictability associated with nonlinear regimes in an atmospheric model

John M. Peters, Sergey Kravtsov, Nicholas T. Schwartz

Research output: Contribution to journalArticlepeer-review

4 Scopus citations

Abstract

Atmospheric regimes are midlatitude flow patterns that persist for periods of time exceeding a few days. Here, the authors analyzed the output of an idealized atmospheric model (QG3) to examine the relationship between regimes and predictability. The regimes were defined as the regions of the QG3 phase subspace characterized by excess persistence probability relative to a benchmark linear empirical model (EMR) for geographically two-dimensional and then zonally averaged flow patterns. The regimes identified correspond to the opposite phases of the Arctic Oscillation (AO + andAO -) and to a more regional pattern reflecting the positive phase of the North Atlantic Oscillation (NAO +). For all of these phase-space regime regions, the leading modes of the QG3 state vector decay to climatology at a slower rate than predicted by the EMR, which contributes to the maintenance of non-Gaussian regime anomalies. Predictable regimes are connected to "regime precursor" regions of the phase space, from which trajectories flow into regime regions following mean phase-space velocities. Packets of trajectories originating from these regions are characterized by anomalously low spreading rates due to a combination of low local stochastic diffusivity and convergence of the nonlinear component of mean phase-space velocities along the trajectory pathways. While unpredictable regimes do have precursor regions, trajectories emanating from these regions are characterized by relatively high spreading rates. The predictable regimes AO + andAO - are insensitive to the metric used to identify the regimes; however, the unpredictable regime NAO + in the 2D space is not directly associated with its zonal-metric counterpart.

Original languageEnglish (US)
Pages (from-to)1137-1154
Number of pages18
JournalJournal of the Atmospheric Sciences
Volume69
Issue number3
DOIs
StatePublished - Mar 2012

All Science Journal Classification (ASJC) codes

  • Atmospheric Science

Fingerprint

Dive into the research topics of 'Predictability associated with nonlinear regimes in an atmospheric model'. Together they form a unique fingerprint.

Cite this