A Dynamical Empirical Orthogonal Function Analysis of the Intraseasonal Disturbances

Abstract

This study investigates the structural properties of the intraseasonal disturbances in the 500-mb flow during a specific winter season (1982/83) and assesses the relevance of normal-mode barotropic instability to the observed intraseasonal disturbances. The leading complex empirical orthogonal function (CEOF1) accounts for 56% of the variance, has a duration of about 40 days, and mainly comprises the EOF1 and EOF4. It is characterized by a strong planetary-scale fluctuation centered over the eastern Pacific region. Its sign changes at the midseason, arising from an eastward shift of the Pacific trough. The CEOF2 explains 22% of the variance with a period of about 20 days and largely consists of the CEOF2 and EOF3. It depicts a rotating dipole structure over the Atlantic region. The CEOF2 is mainly associated with the pronounced blocking episode in February 1983. By applying the Gram-Schmidt method in conjunction with an instability analysis, we formulate a dynamical-EOF analysis. The leading DEOF is found to be a barotropically neutral mode that accounts for 34% of the winter-average spatial variance. It bears some resemblance to the EOF1, implying that a large portion of the intraseaonal variability is barotropically neutral with respect to the seasonal-mean flow. By inference, substantial variability is sustained by forcing. The DEOF2 is a slightly modified unstable mode that accounts for 12% of the variance. It partly resembles the blocking. The results suggest that the atmospheric block under consideration is only partly due to the local barotropic instability of the seasonal-mean flow. A supplementary energetics analysis reveals that the synoptic eddy-forcing mechanism is significant throughout the season, the barotropic instability process occurs intermittently, the dissipation is moderate, and the baroclinic process generally acts against the intraseasonal disturbances. The nonmodal rather the modal form of barotropic instability is probably more important.