Winter Outlook 2005 – 2006

By Eric J. Hörst, Director of the MU Weather Information Center

This is by far the most difficult winter outlook I’ve had to put together. For more than a decade I’ve embraced the challenge of attempting to predict the trends for the upcoming winter season. Last season’s forecast worked out quite well. The call was for a mild December and a late first snow, followed by a change to colder, more volatile conditions with total snowfall between 26 and 36 inches. Well, December 2004 was mild and it didn’t snow until January 17, 2005; following this, there were some wild swings in temperature and several good snowstorms. Seasonal snowfall was 28.5 inches. Looking ahead to the winter of 2005-2006, it’s a much more difficult forecast and my confidence is reduced. Here’s why.

Long-range forecasters leverage a variety of global “signals” and “analogs” in creating a seasonal outlook. Some years these indicators are congruent, leading to a high-confidence seasonal outlook. This year, however, most of these signals and analogs are either contradictory or yield no clue at all. Given this fog of uncertainty, it’s easy for a forecaster to see—and predict—what they want to see. No doubt, a few long-range forecasters are wearing snow-colored glasses, given some of the cold, stormy winter outlooks I’ve heard this fall.

Winter Outlook “Signals”

A wealth of research has revealed the usefulness of numerous global indices in determining climate trends. The best known—and most reliable—is the ENSO cycle in the equatorial Pacific. Depending on the ENSO phase, there are different influences on winter weather in our area. A weak to moderate El Nino often gives birth to frequent snowstorms, whereas a strong El Nino typically signals a mild, wet winter (with rain more common than snow here in Lancaster). The opposite mode, La Nina, overwhelmingly results in a warmer and drier winter in the Eastern U.S. Currently ENSO is neutral, but there are signs that a La Nina may emerge in 2006. The bottom line: Our most reliable signal for making winter outlooks (ENSO), provides no clues this winter.

What about the alphabet soap of other indices--PDO, PNA, NAO, AO, QBO, et al.? Do they reveal any clues? Yes, the Pacific Decadal Oscillation (PDO) is in the cold mode and should continue to maintain an active jet stream off the Pacific into the western Canada and the Northwest coast. Sea-surface temperature trends hint that the cold (negative) PDO may be waning, so this signal could be less of a player later on in the winter season. Still, it seems reasonable that an active Pacific jet would lead to above normal precipitation in the Northwest states and above normal temperatures across parts of the Southwest, southern Rockies, and central to southern plains. Farther east, this jet can also produce spells of warm weather from the Midwest to the East coast; it can also seed storm development for the Northeast U.S.

Another important signal this winter will be the North Atlantic Oscillation (NAO). Unlike ENSO and PDO, which are indices of sea-surface temperature (which changes over the period of months and years), the NAO is an index of atmospheric pressure. A “negative” NAO indicates high-latitude blocking near Greenland and a high likelihood of below normal temperatures (and occasional storminess) in the Northeastern U.S. Conversely, a “positive” NAO often signals ridging along the East coast, a storm track west of the Appalachians, and above normal temperatures in the Eastern U.S. Of course, pressure patterns change over the course of days and weeks, and therefore the NAO often flips several times in a month. However, research has shown that warm water near southwest coast of Greenland and cooler water near the Canadian Maritimes favor a bias toward negative NAO. Although complex and ever-changing, current sea-surface temperature anomalies in the Atlantic resemble the configuration that should favor a more frequent negative NAO this winter. Other indices provide little lead or no signal, in my opinion. I expect the PNA to be highly variable this winter, and I’m not a believer in the validity of the QBO as a signal for winter weather in the East.

Winter Outlook “Analogs”

Analogs can be a powerful lever for opening the door to future weather trends. They can also be a red herring or, worse yet, a pathway to a bad forecast. “Analoging” begins by perusing the data, maps, and impacts of past years in an attempt to cull specific months or seasons that are analogous to current (or recent) patterns, anomalies, and impacts. For example, this September through November has been exceptionally warm. So, looking back to identify similarly warm autumns—and then studying the winter weather that followed—might provide a clue to the evolution of this year’s patterns. Other things one might try to analog (this year) are the record wet October, the near-record warm Summer, or even the active Atlantic hurricane season. Some years you can uncover multiple analogs that provide similar signals—this would provide a high-confidence outlook. This year, however, the analogs are mostly contradictory. It’s an “eye of the beholder” situation, where you can search for and find a single analog that supports your desired outcome. Exercising objectivity, however, one must accept that analoging is not a useful tool when there are contradictory analogs.

Climate Models

Climate models are the latest rage in some camps, and they may very well become a powerful long-range forecasting tool in the future. Right now, however, the skill is marginal and I hesitate to place much weight on models run six months out. Still, NOAA’s Climate Forecast System does hint at a slight cool bias in the East and above normal precipitation in the Northwestern states—so I’ll accept this as a sign that I might be on the right track (or foolish).

Pattern Recognition and Intuition

We have now arrived at the personal attribute that separates the best from the rest in the weather forecasting business. There are really two types of forecasters: 1. Those who “read” the models and then make a forecast to suit (sadly, the common practice and the reason for the ubiquity of mediocre forecasts); 2. Those who have spent tens-of-thousands of hours “living” the weather—developing keen skill at pattern-recognition and the ability to intuit the forecast—and use models as a tool to support their intuitive sense of the weather forecast and changing weather patterns. Such skilled forecasters are rare, but you can spot them by their distinctively unique and accurate forecasts in weather situations that matter, like an approaching hurricane and winter storm. Of course, these pattern recognition skills are extremely useful in preparing a seasonal outlook, too. Hopefully I can add a pinch of said pattern-recognition and intuition to this Winter Outlook!

Winter Outlook

Local weather trends this winter will depend on the influence of the Pacific jet and sign of the NAO. The NAO is the more important signal—but also more transient—and it will ultimate dictate the temperature bias this winter. Based on my analysis of indicators, it’s my sense that the NAO will be biased negative for this winter, especially early on. Therefore, December may be the coldest month of the winter relative to normal. I expect that both January and February will feature cold spells as well, however, occasional mild periods may offset the cold bias late in the winter. The greatest negative (cold) anomaly should be from the Great Lakes to New England, with a seasonal departure from normal of about –3.0 F degrees. Here in southcentral PA, the below normal December and near normal late winter may result in a departure from normal of about –2.0 F degrees. Therefore, this winter may conclude “slightly below normal” overall, but with several notable warm spells.

The greatest precipitation departure from normal should be in the Pacific northwest where up to 150% of normal precipitation may fall. The lee of the Great Lakes and much of the Northeast should also receive above normal precipitation. We’ll need to watch for disturbances born of the Pacific jet that will occasionally trigger cyclogenesis in the Plains, Midwest, or Lower Mississippi valley. Such storms can be prolific precipitation producers in the Northeast, but as always the big question will be precipitation type. At times of negative NAO, these storms will most often pass south of Pennsylvania, and produce significant snowfall. However, during periods of the neutral or positive NAO, these storms will track to our west, leading to rain or snow changing to rain across much of the Northeast. In the end, total precipitation should be 110 to 125% above normal. Seasonal snowfall will be directly related to the dominance of the negative NAO phase. Assuming a season-long bias toward a negative NAO, I would expect snowfall to average a bit above normal as well (say, between 25 and 35 inches here in Lancaster).