Contrary to popular belief, most meteorologists take great pride in being correct and feel deeply troubled when an important forecast misses the mark. Unfortunately, this is what happened to many weather forecasters who dared to make a snow forecast for the winter of 2011/2012 in the U.S. What went wrong? The answer lies in two words that few skiers had heard before that season but are now etched in their memory: Arctic Oscillation.

Before taking a closer look at the Arctic Oscillation and how it can wreak havoc on long-range forecasts, let’s take a look at how meteorologists make seasonal forecasts of three to six months. This type of long-range forecast is far different than the more usual forecasts that predict weather for the next one to seven days. While these short-term forecasts are made by highly sophisticated computer models that usually run on government owned super computers, seasonal forecasts rely much more on using history as a guide to the future.

To create a seasonal forecast, meteorologists look to match current conditions to historical climate patterns like El Nino and La Nina. For example, if the upcoming winter is forecasted to be a La Nina winter, meteorologists will go back in history to look at previous La Nina winters, research how much snow fell during those years, and use that historical data to create a forecast for the upcoming winter. Seems simple, right? It is simple, and often times this method (called “historical analogs”) is too simple to yield a good forecast.

The problem with this historical method is that the amount of snow that falls during a winter is controlled by two types of climate factors, and only one is predictable far in advance. Scientists can predict long-term climate factors like El Nino and La Nina up to 6 months in advance and the phenomena can last 6 to 12 months. However, scientists cannot predict other short term climate factors like the Arctic Oscillation more than a few days in advance, yet these shorter term factors can control snowfall just as strongly as the more predictable, long-term phenomena like El Nino or La Nina.

The Arctic Oscillation is a fancy name for a whirlpool of air around the north pole. When the Arctic Oscillation is said to be in a “positive phase,” this whirlpool of air is stronger. This has the affect of keeping colder air locked up closer to the north pole and does not allow it to move south into the lower 48 states. This keeps most of the snow very far north in Washington State and Alaska with less snow further south.

Now that we’re armed with some background, let’s go back to the maligned snow forecast for the 2011/2012 season and see what went wrong. With a La Nina present, meteorologists expected that season to be similar to the previous 2010/2011 season where many ski areas received plentiful snow. However, the arctic oscillation was strongly positive for much of the early part of the season and never “flipped” over to its negative phase during any part of the season. Since a positive Arctic Oscillation means that the whirlpool of air around the north pole is stronger than normal and doesn’t let colder air move south into the U.S., most of the snow during the 2011/2012 season fell in Alaska (good for them) and in Canada and the Pacific Northwest. Other areas that generally have normal or above normal snow during a La Nina season—like Tahoe, Utah and Colorado—saw far below average snow.

Most seasons, the Arctic Oscillation flips between its positive and negative phases a couple of times and the U.S. sees periods of heavier snow and periods of less snow and warm weather. But during the 2011/2012 season, the Arctic Oscillation never flipped to negative and the strong whirlpool of air up north never let the cold air flow south into the states. This unpredictable climate factor was largely responsible for the inaccurate seasonal snow forecast for 2011/2012.

Short-term weather forecasts of a few days are far more accurate today than 20-30 years ago and are generally very dependable. However, seasonal forecasts for the next three to six months are not very accurate especially when it comes to snow, because short-term climate factors like the Arctic Oscillation exert a great influence on snowfall yet can’t be predicted more than a few days in advance. As a weather forecaster that enjoys making accurate predictions, things like the Arctic Oscillation teach me to stick to short term forecasts and leave the long-range stuff to the Farmer’s Almanac.

Meteorologist Joel Gratz is the creator of and is based in Boulder, Colo.