Background

For ten years at App State Football games, a "vocal" guy has been sitting a few rows behind me yelling in my right ear. Of all the things he yells during a game, my personal favorite is "Watch The Fake!!!!!!!". It's like clock-work regardless of the score, position on the field, or the time on the clock. When the opposing team is lining up to punt or kick a field goal, he yells "Watch The Fake!!!!!!!" (the number of exclamation marks is an indicator of volume). And of course, a couple times each season he is right. That couple of times is enough to convince him that's he's got it figured out, and that a fake kick needs to be foremost in our minds every time the opposing team punts or kicks a field goal.

And so it goes with Winter Forecasting in the Fall every year. Regardless of what has happened leading up to October, the weather drama queens and kings yell "Watch for A Cold/Snowy Winter!!!!!!!". And of course once or twice a decade, they will be right, and that is enough to perpetuate the ritual. It can be fogs in August, squirrels gathering acorns, the number of acorns (and other nuts) on trees, a cool summer, a warm summer, a wet summer, a warm fall, a dry fall, woolly worms, lots of hurricanes, few hurricanes, more foliage, more spiders, where squirrels and hornets build nests, mice invading homes, various behaviors of birds, corn husks, holly and dogwood berries, thickness of animal's fur, big pine cones, corn husks, and I'm sure many more that I have never heard of. Whatever happens each Fall, something can be found as a reason to yell "Watch for A Cold/Snowy Winter!!!!!!!".

Don't get me wrong... I think all the winter forecast lore can be fun and great conversation pieces, as long as we don't take it seriously.

Now the problem for scientists... Seasonal forecasting is an impossible business. Yes, we can see some things coming; however, weather is a very complex system especially when we try to look 3-5 months down the road. Every long-range indicator we have has a thousand caveats. Even the best long-range forecasters are dead wrong sometimes. The most recent example of a great long-range forecaster being dead wrong is the complete bust of the Hurricane Forecast by Dr. Gray (Colorado State) and every other long-range Hurricane Forecast I am aware for this season. A great outcome for long-range forecasters is to be on the right side of average 70% of the time; that's "genius level long-range forecasting".

I'll be honest here... Last fall (and last's winter's forecast was pretty good, not great but pretty good), I declared "never again". I was afraid that even entering into this arena cheapened our brand and sometimes overshadowed what we do best--have fun while being the most reliable forecast for Western North Carolina. But as the time for a Winter Forecast approached, I changed my mind--our readers want a Winter Forecast and there should be a reasonable voice in the midst of a few people yelling "Watch The Fake!!!!!!!".

The Data

ENSO Analysis

The first things we look at when producing almost any long-range forecast are the current state of ENSO (El Nino-Southern Oscillation, a measure of large-scale weather conditions in the Equatorial Pacific) and the forecast for ENSO in the coming months. Currently, the ENSO is neutral (see Figure 1) and is forecast to remain neutral through the coming winter (see Figure 2).

Figure 1: Sea Surface Temperature Anomaly (10/17/2013)

Figure 1 shows sea surface temperatures as of October 17, 2013. Blues are cooler than normal water; yellows-reds are warmer than normal waters. The region of interest for ENSO is the Equatorial Pacific from Chili to northeast of Australia. An average significantly warmer than normal indicates El Nino. An average significantly cooler than normal indicates La Nina. While there are areas of both warm and cool water shown in Figure 1, the overall average in the Equatorial Pacific (the ENSO region) is currently near normal.

Figure 2: Forecast for ENSO in 3-month Intervals

Figure 2 shows computer model forecasts for ENSO. Note the "DJF" along the horizontal axis; that's the forecast average for December-January-February. Only 7 of these have a value for DJF above +0.5 (El Nino) or below -0.5 (La Nina) and the average of all these models results is very near 0. It's about as neutral a forecast as one could ever have.

So, our winter forecast for 2013-2014 is based on neutral ENSO conditions. Figure 3 shows total winter snow accumulation in Boone, NC classified by ENSO (Strong El Nino, Moderate El Nino, Neutral, Moderate La Nina, and Strong La Nina).

Figure 3: Total Winter Snow in Boone NC Classified by ENSO

Note that Strong El Nino Winters have slightly less snow than average. Moderate El Nino winters have by far the most snow. Strong La Nina winters have much less snow. Neutral, and Moderate La Nina winters have near average snow.

Note also how much variation exists from year to year (indicated by the large Standard Deviation). Of the five ENSO classifications, Moderate El Nino and Neutral years have the most variation from year to year in snow totals. Exhibit A in illustrating this variation in Neutral years is that both 1959-1960 (the most total snow ever recorded) and 2001-2002 (the least total snow ever recorded) were Neutral years.

A Neutral ENSO is a "weaker signal" meaning that other factors (and more difficult to forecast from 3-5 months out) will have greater influence on cold and snow. Another primary factor in cold/snow in the Eastern U. S. is the level of "high latitude blocking" as indicated by the Arctic Oscillation (AO) and North Atlantic Oscillation (NAO). These indicators have much shorter cycles than ENSO and are much harder to forecast 3-5 months out. We do look at computer model forecasts for the AO and NAO for 30-45 days out as a good indicator of temperature and snow potential in the next 4-6 weeks. As of mid-October, the AO/NAO forecast hint at colder weather toward the end of November and beginning of December. But beyond 45 days, these forecasts are of little use.

So the starting point for our snow forecast will be a snow total near average since Neutral ENSO winters have an average near the 54-year average; however, we have less confidence in this forecast since Neutral winters have more variability in snow totals than other classifications.

Recent Winter Trends

Figure 4 shows the 10-year average winter snowfall for Boone NC for each year since 1968.

Figure 4: 10-year Snowfall Average for Boone NC

This graph clearly shows the decrease in yearly snowfall in Boone NC between 1985 and 2008. That 10-year average had a significant increase in the last four years because 4 of the last 5 years had more snow than the 10-year average, including 2009-10 that was the third snowiest winter ever recorded. Note that only 6 of the 23 winters since 1990 had more snow than the 54-year average.

So, based on snowfall in the last 23 winters being well below the 54-year average, we will hedge our forecast below the 54-average.

Arctic Ice

Figure 5 shows current ice cover in the Arctic relative to average and last year. (2012 had the least Arctic ice coverage ever recorded.)

Figure 5: Extent of Arctic Ice

In previous years, we have used the extent of Arctic Ice as a factor in the forecast; however, we no longer think the implications of Arctic Ice and winters in the Eastern U. S. are well understood. Overall one would think that less ice would contribute to a warmer atmosphere (and vice versa) since ice reflects more solar radiation (compared to liquid water that absorbs more solar radiation). However, some scientists believe that this phenomenon may have contributed to high latitude blocking and a persistent negative NAO in the winter of 2009-10 resulting in a persistent trough in the Eastern U. S. and correspondingly lots of snow. So, we will not use the ice extent in the Winter Forecast this year.

Long-Range Computer Modeling

A survey of long-range computer modeling confirms our thinking of near-average overall but low confidence. There's a hint of warmer in some months; however, the signals are too weak to be of use in the forecast (other than confirm what we already thought).

The Fearless Forecast

So, with all the signals weak and inconclusive, what does a forecaster do? We'll do the same thing we do when driving down the Blue Ridge Parkway in fog--we maintain focus on the center-line and keep the left wheel as close as possible to that center-line.

Bottom Line for the Winter 2013-2014 RaysWeather.Com Fearless Forecast:

• Slightly less snow than the 54-year average
• Slightly warmer than normal temperatures

Here is our forecast for many locations in Western North Carolina reflecting that strategy:

Location	Expected Total Snow/Ice for Winter 2013-2014
Asheville	13"
Banner Elk	43"
Beech Mountain	100"
Boone	38"
Hickory	6"
Jefferson and West Jefferson	26"
Lenoir	7"
Morganton	8"
Mount Airy	10"
Sparta	23"
Spruce Pine	23"
Sugar Mountain	90"
Waynesville	13"
Wilkesboro	10"

With that, we return to our regularly scheduled daily forecasting duties. We'll keep you informed with the most reliable forecasts for Western North Carolina all winter long.