Robustness of Low-Frequency Circulation Patterns Derived from EOF and Rotated EOF Analyses

Open Access
Publisher Website
Google Scholar
Abstract
The robustness of low-frequency circulation patterns defined by unrotated and rotated empirical orthogonal functions (E0Fs) are compared based on the Northern Hemisphere 10-day low-pass filtered wintertime 500-hPa height field. The Pacific/North American pattern and the North Atlantic Oscillation are the most prominent modes of low-frequency variability in the data. The reproducibility of the spatial patterns derived from EOF and rotated EOF analysis is assessed by repeating the analysis on 50 subsets of the data, each comprised of the maps belonging to 22 randomly selected winters from the 44 winters in the total record. The results indicate that rotated spatial patterns are less sensitive to sampling fluctuations than their unrotated counterparts. Abstract The robustness of low-frequency circulation patterns defined by unrotated and rotated empirical orthogonal functions (E0Fs) are compared based on the Northern Hemisphere 10-day low-pass filtered wintertime 500-hPa height field. The Pacific/North American pattern and the North Atlantic Oscillation are the most prominent modes of low-frequency variability in the data. The reproducibility of the spatial patterns derived from EOF and rotated EOF analysis is assessed by repeating the analysis on 50 subsets of the data, each comprised of the maps belonging to 22 randomly selected winters from the 44 winters in the total record. The results indicate that rotated spatial patterns are less sensitive to sampling fluctuations than their unrotated counterparts.