Winter 2009 forecast for North America
Summary: In the winter of 2009, colder than normal temperatures are forecast for the west coast of North America (from Alaska to the U.S. Pacific Northwest) and the eastern United States, particularly in the Midwest. An anomalously cold spell is expected for the Southeast in January 2008. Warmer than normal temperatures are likely in the southeastern United States and northeastern Canada.

Fig. 1. Forecast of surface air temperature anomalies in North America in the winter (DJF) of 2009. 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
In the absence of a pronounced ENSO event, there is a great deal of uncertainty about the distribution of temperature anomalies during the winter of 2009. As in the previous winter, the PDO index is forecast to be negative (see the forecast for the North Pacific). Therefore some aspects of the forthcoming winter may be similar to those of the winter of 2008. In particular, colder than normal temperatures are expected for the west coast of North America, from Alaska south to the US Pacific Northwest.
The Midwestern United States will likely to see frequent outbreaks of cold Canadian air, which was also a feature of the winter of 2008. The cold outbreaks during that winter were associated primarily with the La Niño event. In the forthcoming winter, however, the outbreaks are suggested by rules 475 and 522. The first of these rules links winter temperatures in the Midwest with the Arctic Oscillation (AO) index, (see the forecast for the North Atlantic), and the second one with the West Pacific (WP) index (see the forecast for the North Pacific). Both of these indices are forecast to be negative.
When the PDO index is negative, the Pacific-North American (PNA) index tends to be negative as well. The forecast of the negative PNA index is also supported by rule 199. However, rule 200, as well as the extrapolation of year-to-year changes in the PNA index suggest that the index may be positive. After the confidence factors are combined, the PNA index is forecast to be near zero or slightly positive for the whole winter. The PNA is more likely to be negative in the early winter (Nov-Dec) and positive from January through March. This forecast is also consistent with rule 227 that links the PNA index with sea-level pressure (SLP) in the Gulf of Alaska. The letter is forecast to be negative in January 2009 based on solar activity and QBO (rule 198).
Mean winter temperatures in the Southeastern United States are forecast to be below their 1971-2000 averages, mainly due to a negative NAO pattern (see the forecast for the North Atlantic). However, it should not be extremely cold there, because of the influence of a negative PDO pattern. The coldest spell will likely to occur in January when a negative NAO pattern will combine with a positive PNA pattern.
Overall, winter temperatures are expected to be below normal over most of the United States, except the Southwest. This forecast is in sharp contrast with the winter 2009 outlook from the Climate Prediction Center. According to the latter, the area of enhanced chances of above normal temperatures includes most of the U.S., except the area west of the Rocky Mountains. The forecast by the CFS model is more in line with our forecast.
Regarding the recommendations for trading at Weatherbill.com, the best bets are for below normal temperatures in the Pacific Northwest in November-December and February-March; below normal temperatures in the Midwest for the entire season; and below normal temperatures in the Southeast in January.