Winter 2008 forecast for North Atlantic - update

Summary: In the winter of 2008, the North Atlantic Oscillation (NAO) index is forecast to be near zero or slightly positive. Above normal sea surface temperatures (SSTs) will dominate, consistent with the continuing positive phase of the Atlantic Multidecadal Oscillation (AMO). The basin-wide SST anomaly, however, will be lower than during the previous winter. The highest CFs are for the positive SAT anomalies in the central North Atlantic and off the southeastern United States. It is likely that SAT anomalies will also be positive in both the Northwest Atlantic and the Norwegian/Barents Seas, which is a characteristic feature of SST anomaly distribution since 1997. Negative SST anomalies are expected in the East Greenland Current, off Newfoundland and along the West African coast.

Fig. 1. Forecast of sea surface temperature anomalies in the North Atlantic in the winter (DJF) of 2008.  The blue (red) contours and the letter C (W) indicate colder (warmer) than normal temperatures. CL means near normal. The base period for calculating anomalies is 1971-2000. The numbers at the contours are confidence factors of the forecast.

Discussion

The North Atlantic Oscillation (NAO) index is expected to be near zero or slightly positive. This is because the effects from different predictors almost counter-balance each other.

One piece of evidence in favor of a negative phase of the NAO comes from the analysis of the North Atlantic SST anomalies in May (rule 22). This predictor is used by the UK Met Office in their statistical predictions of the winter NAO. The forecasting procedure was developed by Rodwell et al. (1999) and Rodwell and Folland (2002). The projection of May 2007 SST anomalies onto the predictor pattern produces a slightly negative (-0.05) value of the NAO, with a standard error of ±1.0. Other authors use June-October or July-September SST anomalies instead. In either case, the analysis of SST anomaly patterns is not very conclusive, although a positive SST anomaly south of Greenland in recent months is more suggestive of a negative phase of the NAO.

Summer surface air temperature (SAT) anomaly pattern in the subpolar region is also used to predict the NAO index in the following winter (rule 24). In June-July 2007, the SAT subpolar index was one standard deviation below normal, which indicates that the NAO index may decline. The negative value on the subpolar index is largely due to a record warm summer in southern Greenland. 

Many climatic indices, including the NAO, have a 60-80-year component in their variability. Currently, the NAO is in the declining phase of this cycle that reached its maximum in the 1990s. Therefore, negative values of the NAO are getting more and more likely. Nevertheless, the NAO index in the winter of 2008 is expected to be slightly positive due to the following reasons. 

A major contributor into the confidence factor for the positive phase of the NAO is a sea level pressure index in the North Pacific. The correlation between the Pacific index and the NAO is statistically significant for the entire period of observations since 1951, but is particularly strong in recent decades reaching 0.67 for the period 1972-2007 (rule 38). The 2007 value of the Pacific index was second largest (after the 1980 value) on record, thus strongly suggesting a positive NAO index in 2008.

It was also shown that the winter NAO index is linked to summer snow cover anomalies in Eurasia (rule 23) and Northern Hemisphere (rule 25). In the summer of 2007, the snow cover both in Eurasia and the Northern Hemisphere as a whole was well below normal, suggesting a possibility of the positive NAO index in the forthcoming winter. 

The SLP anomaly in the Azores high in 2008 is expected to be close to normal or slightly positive. It is also likely that it be shifted westward, so that its influence on SAT in the southeastern United States will increase (see the forecast for North America). Mild winters in the southeastern US are consistent with the positive NAO phase. 

Despite the forecast of a positive NAO index, SST anomalies in the Baffin Bay/Davis Strait area are expected to be positive. Ice cover in Baffin Bay is closely linked (rule 189) to the Greenland blocking index, which in turn is correlated (rule 53) with solar activity. The latter suggests that the Greenland blocking index is likely to be positive, which is often accompanied by below normal ice cover. 

It appears that the winter of 2008 will be another one with positive SST anomalies in both the Northwest Atlantic and the Norwegian/Barents Seas. Typically these areas exhibit an opposition in SST/SAT fluctuations, which can be used as another NAO index. Indeed, the correlation coefficient between the difference in winter SAT anomalies in Oslo, Norway, minus Jakobshaven, West Greenland, and the SLP-based NAO index is 0.67 for the period 1981-2007. During the periods of warm climate, as in the 1930s, the frequency of the situations when SAT anomalies are positive in both places increases (Fig. 2). The pattern became more frequent again since 1988 and persistent since 1997. In 2007, the sum of SAT anomalies in Oslo and Jakobshaven reached the highest value on record (Fig. 2). 

The increased frequency of this pattern also corresponds to the positive phase of the Atlantic Multidecadal Oscillation (AMO). It should be noted that the North Atlantic cooled down substantially (in terms of monthly SST anomalies) from November 2006 to May 2007. Since then, however, the basin-wide SST anomaly is on the rise again. In the winter of 2008, the North Atlantic is expected to be warmer than the average for 1951-2000, but not as warm as during the previous winter.

Fig. 2. A sum of normalized (by standard deviations) surface air temperature anomalies in Oslo, Norway, and Jakobshaven, West Greenland, 1881-2007.