EOF and SVD Analyses of the Low-Frequency Variability of the Barotropic Component of the Atmosphere

Abstract

In this study, we have investigated the dominant horizontal patterns of the barotropic component of the atmosphere, the external forcing to it, and the nonlinear scale interactions, using the EOF and SVD analyses for 50 years of DJF mean NCEP/NCAR reanalysis. According to the EOF analysis, the horizontal pattern of the EOF-1 of the atmosphere represents the Arctic Oscillation (AO), and the EOF-2 represents the PNA teleconnection pattern. The SVD-1 for the barotropic component and the external forcing represents the structure of AO with the external forcing characterized by topographic forcing. The energy balance for the SVD patterns, however, shows that the external forcing of the SVD-1 damps the AO-like anomaly of the barotropic component. The SVD-2 represents the PNA, and the time series of the coefficient indicates a link with El Niñ o/La Niñ a. The external forcing of the SVD-2 appears also to damp the PNA-like anomaly. Furthermore, the SVD-1 for the barotropic component and the nonlinear scale interactions represents the AO-like anomaly with the large nonlinear scale interactions over half of the Northern Hemisphere. The energy balance for this SVD pattern shows that the nonlinear scale interactions excite the AO-like anomaly. Similarly, the SVD-2 shows the energy increase by the nonlinear scale interactions the same as SVD-1. The nonlinear scale interactions are further decomposed in linear interactions with the climate basic state and the reminder. We find by the energetics analysis that the AO-like anomaly is mostly excited by the interactions with the basic state.It is concluded from the SVD analyses that the extreme events associated with low-frequency variabilities are not induced by the low-frequency forcing, but mostly by the energy supply from the climate basic state.

Ensemble‐based singular value decomposition analysis to clarify the causes of heavy rainfall

“…These multivariate statistical analyses use datasets from as many analysis times as the statistical population. At the time scales governed by the datasets, the synoptic‐scale climatological features of the atmospheric and oceanic fields (El Niño–Southern Oscillation [ENSO] and Arctic Oscillation [AO]) are discerned using SVD analysis or CPCA (Wallace et al ., 1992; Koide and Kodera, 1999; Matsueda and Tanaka, 2005). SVD analysis can also be applied to clarify the climatological relationship between mesoscale features of precipitation and synoptic‐scale atmospheric states (e.g., Kang et al ., 2007; Chu et al ., 2008).…”