Seasonal to Interannual Forecasts
Extreme Weather: Background
The wintertime Pacific/North America (PNA) pattern is the dominant mode of seasonal/interannual variability over the Pacific/North America sector and the most predictable northern hemisphere circulation pattern at lead-times of 6-10 days (Renwick and Wallace 1995).
Using daily weather station data for 1948-1998 (October-March only), we have demonstrated strong associations between the statistics of extreme daily weather events (those days in which the parameter of interest is more than 1.5 standard deviations from the daily mean value) and the phase of the daily PNA index in coastal Alaska, the US Pacific Northwest, around the Great Lakes region, and in the Southeastern U.S. Parts of the Southeast, for example, have experienced extreme cold temperatures 4-8 times more frequently during positive PNA days, while parts of the western U.S. and coastal Alaska have experienced 8-20 times more extreme cold temperatures during negative PNA days (Figure 1).
Similar analyses demonstrate the relationship between the PNA index and the relative frequencies of extreme daily precipitation, freezing temperature days, snow days, maximum temperatures, and surface wind gusts. This work shows that existing operational PNA forecasts (from NCEP) can be used to generate skilled extreme event risk forecasts for select locations at lead times up to two weeks.
click image to enlarge
Figure 1 The statistics of extremely cold temperatures in relation to the daily PNA index. The red and blue dots indicate the relative frequency of extremely cold daily temperature minimums (more than 1.5 standard deviations from the daily mean) recorded during positive versus negative PNA days. Red (blue) wind barbs are anomalous surface winds on negative (positive) PNA days.