NAO, PNA and zonal winds

Maarten H. P. Ambaum and Brian J. Hoskins
University of Reading, UK

 

Definitions of the PNA and NAO

PNA

 

 

 

NAO

The PNA is associated with a Rossby wave pattern with centres of action over the Pacific and over N. America. The NAO is associated with a north-south orientated dipolar structure in the pressure field over the Atlantic. A useful definition of the two patterns is in terms of EOFs of the streamfunction at 850 hPa. The first EOF represents the PNA and the second EOF the NAO. Both patterns correlate well with more traditional definitions of these patterns, such as teleconnections.

First two EOFs for the DJFM monthly mean streamfunction at 850 hPa. These EOFs explain 29% and 14% of the variance respectively. The contour interval is 5x105 m2s-1.

 

Influence on 250 hPa zonal winds

PNA

 

 

 

 

 

NAO

The effects of the PNA and NAO on the zonal winds at 250 hPa are remarkably similar even though the index timeseries here are orthogonal by definition. Both patterns correspond to a strengthening of the subtropical jet over the Atlantic and Arabia and a strengthening and Northward shift of the main jet off the east coast of N. America. The PNA pattern is also associated with a weakening and broadening of the Pacific jet, whereas the NAO only weakly influences the Pacific region.

Monthly mean zonal wind field regressed on the normalised timeseries of the PNA (top panel) and NAO (lower panel). The contour interval is 1 ms-1.

 

Influence of PNA and NAO on zonally averaged winds

Climatologies of zonally averaged zonal winds for the Atlantic sector (60W-0W, top four panels) and the Pacific sector (150E-120W, bottom two panels) for different values of the NAO and PNA indices.

Increasing either the NAO or PNA index leads to similar effects in the Atlantic.

Increasing these indices strengthens and splits the jets in theAtlantic.

The Pacific jet weakens and broadens on increasing PNA index.

The NAO does not affect the Pacific jet.

The polar night jet strengthens on increasing NAO index.

The PNA index affects the polar night jet in the Pacific sector only.

 

Relationship ith the Arctic Oscillation

The NAO and PNA patterns both represent a tendency for the zonal winds at 35N and 55N to anticorrelate. This tendency for anticorrelation appears over both the Atlantic and Pacific basins and the Arctic Oscillation is a reflection of this.

First EOF of the DJFM monthly-mean mean sea-level pressure. It explains 25% of the total variance. The contour interval is 0.5 hPa

The presence of centres of action over both the Atlantic and the Pacific basins is not a consequence of a correlation between lower tropospheric fields in the Atlantic and the Pacific.
The covariance matrix for the mean sea-level pressure in the three centres of action is:

where the rows/columns represent the Azores, Arctic and Pacific centres of action respectively. The first eigenvector, explaining 59% of the variance in this three component system is (0.38,-0.86,0.33), which corresponds to the Arctic Oscillation. The correlations between the Pacific and the other centres of action are however not significantly different from zero.

 

Relationship with annular modes: some questions

What do the words `annular' and `mode' mean?

The zonal mean zonal winds at 35N and 55N tend to anticorrelate in both ocean basins in association with the PNA and NAO patterns. Is this a coincidence or is it a sign of annular behaviour?

Zonally symmetric anomalies affect climatological features at different latitudes in different zonal sectors differently. Should annular behaviour be zonally organised along latitude circles or along climatological streamlines?

Conclusions:

The PNA and NAO patterns both result in anticorrelation of the zonal mean zonal winds at 35N and 55N.

The effects of the PNA and NAO patterns in the Atlantic sector on the zonal mean zonal winds are very similar.

The Arctic Oscillation reflects the tendency in both ocean basins for anticorrelation between the zonal mean zonal winds at 35N and 55N.

The mean sea-level pressure in the Pacific is not significantly correlated with the mean sea-level pressure over Iceland or the Azores, so the Arctic Oscillation cannot be interpreted as a teleconnection.

Zonally symmetric anomalies act on zonally asymmetric climatology leading to significant differences in behaviour in the two ocean basins.

Reference

M.H.P. Ambaum, B.J. Hoskins and D.B. Stephenson (2000): ArcticOscillation or North Atlantic Oscillation. J. Climate (in review)
A preprint may be obtained from: http://www.met.rdg.ac.uk/~dynamic/maarten/research

 

This work was done as part of the EC-funded project STOEC. The STOEC webpage can be found on http://www.met.rdg.ac.uk/cag