Technique for Identifying the Northern Hemisphere Teleconnection Patterns.
The procedure used to identify the Northern Hemisphere teleconnection patterns and indices is the Rotated Principal Component Analysis --RPCA (Barnston and Livezey 1987, Mon. Wea. Rev., 115, 1083-1126). This procedure isolates the primary teleconnection patterns for all months and allows time series of the patterns to be constructed. For our monitoring purposes, we apply the RPCA technique to monthly mean standardized 500-mb height anomalies obtained from the CDAS in the analysis region 20°N-90°N between January 1950 and December 2000. The anomalies are standardized by the 1950-2000 base period monthly means and standard deviations.
For each of the twelve calendar months, the ten leading unrotated EOFs are first determined from the standardized monthly height anomaly fields in the three-month period centered on that month: [i.e., The July patterns are calculated based on the June through August monthly standardized anomaly fields]. A Varimax rotation is then applied to these ten leading un-rotated modes, yielding the ten leading rotated modes and their time series for that calendar month. Therefore, these ten leading rotated modes for each calendar month are based on 153 (51 x 3) monthly standardized anomaly maps. Click here for additional discussion on how the teleconnection indices are calculated.
An examination of all twelve sets of rotated modes revealed ten dominant teleconnection patterns, of which eight to nine appear in each of the twelve calendar months. These patterns are referred to as the North Atlantic Oscillation, the Pacific/North American teleconnection pattern, the East Atlantic pattern, the West Pacific pattern, the East Pacific – North Pacific pattern, the East Atlantic/ Western Russia pattern, the Tropical/ Northern Hemisphere pattern, the Polar-Eurasian pattern, the Scandinavia pattern, and the Pacific Transition pattern.
This analysis accounts for variability in the structure and amplitude of the teleconnection patterns associated with the annual cycle of the extratropical atmospheric circulation. It also allows for better continuity of the time series from one month to the next, than if the patterns were calculated based on the data for each month independently. Finally, the analysis allows for more robust results due to the increased number of fields used to calculate the modes for each calendar month. The RPCA procedure is superior to grid-point-based analyses, typically determined from one-point correlation maps, in that the teleconnection patterns in the RPCA approach are identified based on the entire flow field, and not just from height anomalies at select locations.
The teleconnection patterns are displayed with the positive phase shown in all maps. Note that the sign of the plotted values is reversed for the negative phase of the pattern. For the calendar months shown, the plotted values represent the temporal correlation during 1950-2000 between the index time series valid for that calendar month and the standardized monthly anomalies in the three-month period centered on that month. For example, the plotted values at each point for the January NAO pattern represent the temporal correlation between the January NAO index values and the standardized monthly anomalies for all December, January, and February months during the period 1950-2000.
Historical Archive of all Indices: Monthly Tabulated Indices for all teleconnection pattern amplitudes dating back to 1950.
Monthly Tabulated Indices: Last 12 months of indices for selected teleconnection patterns, as appears in the Climate Diagnostics Bulletin.
Monthly Plotted Time Series: Time series of pattern amplitudes for the last few years for selected teleconnection patterns, from the Climate Diagnostics Bulletin.
Back to contents
Link to The Climate Diagnostics Bulletin
Link to Annual Climate Assessment
Link to Special Climate Summaries