Skip navigation.
Home resources

What is the North Atlantic Oscillation?

(This note prepared with the help of: Dr. Rob Wilby, Department of Geography, University of Derby.)

In very crude terms, it is possible to visualise the mean sea level pressure patterns affecting the north-east Atlantic as varying (or 'oscillating') between two extremes: At one extreme is a minimally perturbed westerly type, with disturbances rattling swiftly across the Atlantic, hurried along by very strong mid / upper tropospheric winds. At the other extreme lies a weak, perhaps ill-defined pressure pattern, but with a strong tendency for stagnation of weather types over and downwind of the north-east Atlantic (a persistent 'blocked' type). [ use of the word 'oscillation' in the popular mind implies some sort of regularity, which in reality is not observed, at least not on short time scales. ]

One method is to use a measure of the 500 mbar strength between defined latitudes: By taking the difference between mean 500 mbar contour heights between latitude 35 and 60 North, this simple method yields high numbers ( a high zonal index ), for strong westerly types, and low numbers ( a low zonal index ), for weak westerly, or blocked types. Another method would be to categorise circulation types using, for example, Lamb's Weather Classification.

However, a simple measure, using observed msl pressure differences from long-term 'normals', can be employed, and can of course be extended back to times well before upper air information became routinely available. Upwind of the British Isles / NW Europe, stations in Iceland and the Azores (and / or Gibraltar) are used, by convention, to define the North Atlantic Oscillation Index (NAOI).

The method of calculation means that lower than normal mslp over the Iceland region and / or higher than normal mslp in the Azores / Gibraltar zone gives rise to a +ve NAOI. The converse situation gives rise to -ve NAOI values. The Index is of most use during the winter, when highly positive values are associated with warmer, wetter, windier winter seasons, especially over NW Europe.

In a record since 1823, the following major 'divisions' can be identified in the winter (December, January, February) season:

  • 1830's to ~1900: +ve overall, but with fairly frequent -ve seasons.
  • 1900 to first half 1930's: strongly +ve (similar to modern times) with only 1 marked -ve season (1916/17).
  • second half 1930's to ~1970: a gentle decline overall, with a series of strongly -ve phases; includes the severe events of 1947, 1962/63 & the second most -ve winter of 1968/69.
  • from 1970: irregular, then from 1980 (@/see note below) strong rise to current generally +ve phase; few strongly -ve winters after 1980, BUT the most -ve winter in the entire series occurs in 1995/96 and it is interesting to note that the long-term 'moving average' has shown a downward turn in recent years.

( @ Note: The strongly positive NAOI since the 1980s has resulted in a higher frequency of unusually wet and mild winter conditions over most of NW Europe and Scandinavia during this period, with concomitant changes in regional runoff.)

to see data relating to the NAOI go to: and other interesting links therefrom.