Every month so far this year has been warmer than average, and so have the past four winters. While a shot of abnormally chilly air is currently upon us, we don’t necessarily see this as a sign of things to come.

We’ve pored over the data, analyzed different climate patterns and expect this winter to be pretty similar to the past several, with somewhat above-average temperatures overall. But even generally mild winters can have harsh stretches, and we expect that this one will be no different.

As for snowfall, we don’t see strong signals for blockbuster amounts or a snow shutout, either, but rather typical totals for the region given recent history — close to or a little below average.

Starting mild but trending colder

While the winter may be somewhat warmer than normal overall, you may notice it becoming more intense with time. We’re forecasting temperatures in December and January to be somewhat warmer than average, but near average in February.

We don’t make an official temperature forecast for March, but signals suggest it could be on the cold side. (Our outlook covers the meteorological winter months of December through February.)

Snow may also pick up in winter’s second half

As a more persistent wintry pattern tries to emerge, we predict that February will be the snowiest month, perhaps followed by March.

However, unless we get a big storm, total amounts will probably fall short of average. The odds of a big storm (of 12 inches or more) are slightly below average this year.

This doesn’t mean we won’t see disruptive events and challenging forecasts. We almost always do.

Overall, we’re forecasting five to seven accumulating snow events in the immediate D.C. metro area, with a couple more in our colder suburbs to the north and west. This doesn’t include dustings or ice events, and we are likely to have some of those, too.

Winter snowfall totals, while near to a little below average, should be around or slightly above the median amount (around 11 inches at Reagan National Airport). This means amounts ranging from 10 to 18 inches inside the Beltway and up to 25 inches or so in our normally colder areas to the north and west.

Outlook details


Overall, we expect temperatures for December through February (relative to the 1981-2010 average) to finish slightly to somewhat above average (around two degrees above average).

Here is the month-by-month breakdown:

  • December: Three degrees warmer than average.
  • January: Two to three degrees warmer than average.
  • February: Around average.
  • March: We don’t make specific predictions for March, but we do think it will be colder than average.


Our snowfall projection covers November through April (1981-2010 statistics in parentheses). Overall, we expect slightly below-average snowfall, but near or slightly exceeding the median.

Here are the predicted snow amounts, by location:

  • Alexandria, Arlington and Prince George’s counties and the District: 10-16 inches
  • Fairfax, Loudoun, Montgomery counties: 15-25 inches
  • Reagan National Airport (DCA): 8-14 inches (compared with a 15.4-inch average, 11-inch median)
  • Dulles International Airport (IAD): 16-24 inches (compared with a 22.0-inch average, 16-inch median)
  • Baltimore-Washington International Marshall Airport (BWI): 15-20 inches (compared with a 20.1-inch average, 15-inch median)


  • Although advances have been made in seasonal forecasting, there is still a great deal of uncertainty and limited skill in developing these outlooks. These remain low-confidence forecasts, and the forecast for this winter is a bit less confident than usual.
  • Note that monthly temperature predictions are less reliable than for the whole season. A cold or warm pattern lingering a week too long or ending a week early can greatly alter a monthly average. Furthermore, it takes only one big snowstorm for us to reach or exceed our seasonal average.

Answers to questions you may have

What are other media outlets predicting?

Several local news organizations have released their winter outlooks, and their predictions do not differ much from our general ideas, but tend to be a bit snowier:

  • NBC4 is calling for 18 to 25 inches in the city (compared with our 10 to 16 inches) and similar temperatures, with December and January on the milder side before trending colder in February and March.
  • FOX5 is calling for 15 to 25 inches of snow and near-average temperatures.
  • WUSA9′s forecast is the closest to ours, calling for 10 to 20 inches of snow and somewhat above-average temperatures.
  • ABC7 has yet to release its outlook.
  • The National Weather Service does not issue a snowfall forecast, but its winter outlook leans toward above-normal temperatures and precipitation for the region.

What is your long-term track record with these winter outlooks?

We have been doing winter outlooks since 2005-2006 and have evaluated ourselves after the fact for the past 12 winters. We’ve generally been in the ballpark, giving ourselves an average grade of around B-, although we’ve had notable triumphs and misses.

Our outlook for last winter was pretty good, as we correctly called for slightly above-average snowfall. It was also solid the year before, when we rightly called for below-average snowfall. Predicting the month-to-month specifics was less successful both years, but this is always the most challenging aspect of the outlook.

Since initiating these outlooks, our best winter forecast preceded the record-breaking Snowmageddon winter of 2009-2010, when we said: “Overall, we find chances for a large snowstorm of 8-12 inches or more are much higher than normal this coming winter.” Our outlook for the winter of 2014-2015 was also quite successful, as we correctly called for it to be cold with somewhat above-normal snow.

Our worst outlooks were for the winters of 2011-2012 and 2013-2014. In 2011-2012, we called for near-normal temperatures and it was five degrees warmer than average. Several winters ago (2013-2014), we called for a warm winter with slightly below-average snowfall, and it was cold, with snow that totaled more than twice the average.

Aren’t weather forecasts only reliable out to about eight to 10 days?

It is true that there is no skill in predicting specific conditions, such as the exact temperature and amount of rain or snow for a given day, more than eight to 10 days into the future. However, seasonal forecasting has advanced to the point that we can make educated guesses on the overall tendency of conditions, such as how temperatures and snowfall will compare to average over a month or period of several months. Because of the uncertainty involved, we give ranges and attempt to be as transparent as possible in conveying that these outlooks are indeed low confidence.


Below are some, though not all, of the factors that we considered in determining conditions for this coming winter.

No single factor tells the whole story, nor are the correlations between past conditions and future conditions — which we used to inform the outlook — always strong. But we have chosen factors that in the past have proved to have at least some predictive value. And when considered collectively, they help paint a picture of what we believe is most likely to happen this winter.

No two winters are alike, but we expect this winter to share some similarities with the winter of 1959-60. This analog helped to very loosely form the basis of our temperature and snow predictions, because the weather then had some similarities to the factors below.

Tropical Pacific Ocean

We are experiencing neither an El Niño nor La Niña event, but the neutral phase in between that is leaning slightly more toward an El Niño. We believe these conditions will persist throughout the winter, but the chances of an official weak El Niño event are not insignificant (perhaps 25 percent chance).

Neutral events don’t have a blueprint in terms of their effects on Mid-Atlantic winter weather, but we think this one will take on some of the characteristics of a weak El Niño. One thing many weak El Niño events have in common are periods where the subtropical jet stream can make its presence felt.

An active subtropical jet stream is indicated by storm systems hitting the California coast and traversing across the southern half of the United States. These storms can move up the East Coast and produce snow.

We expect the subtropical jet stream is most likely to flex its muscles in the second half of winter, but the prevalent, default pattern throughout winter will be a more active northern branch of the jet stream.

When the northern stream is dominant, the prevailing storm track is to our west. When storms track to our west, they tend to draw relatively mild and sometimes dry air over our region. When it does snow in these situations, a flip to sleet or rain is common.

North Pacific Ocean

The Pacific Decadal Oscillation (PDO) is a measurement of the intensity and location of sea-surface temperature differences from average in the North Pacific. When it is strongly positive, it often correlates with a cold and stormy pattern for the Mid-Atlantic as the jet stream lifts up toward Alaska in the West (delivering warm, dry weather to California) and dips across eastern North America. When it is sharply negative, conditions often, but not always, trend warm and dry in the Mid-Atlantic.

After a multiyear persistent positive PDO period from 2014-2017, we are now experiencing more of a neutral period.

We think the PDO will hover close to neutral on balance for the winter, although it could shift more positive if El Niño develops, which would increase the chance for cold, snowy periods.

Arctic Oscillation (AO) and North Atlantic Oscillation (NAO)

The AO is a measurement of surface air pressure difference from normal at the high latitudes over and north of Greenland. Pressures lower than normal indicate the positive phase, and pressures higher than normal, the negative phase.

During the positive phase of the AO, cold air tends to be locked up over the Arctic by a strong polar vortex, and the mid-latitudes tend to be milder than average. During the negative phase of the AO, the polar vortex becomes disturbed, and cold air outbreaks become more likely over the mid-latitudes, including the United States.

The AO’s cousin, the NAO is technically a measurement of the differences in air pressure over the North Atlantic Ocean. It is often indicated by either an upper-level low-pressure area (positive phase) or upper-level high-pressure area (negative phase) over or near Greenland. Often, though not always, the AO and NAO share the same phase, especially when averaged over the course of the winter.

A negative AO in the winter months often correlates with a cold pattern in our region and supports winter storms when other factors align with it, particularly when we have a negative NAO as well. This was the major factor in our historically snowy winter of 2009-2010.

High pressure over Greenland or high-latitude blocking helps push the storm track farther south and east, often creating storm tracks that are cold and snowy for our region. Although a negative AO and NAO combination far from guarantees a cold and snowy period, our chances of a meaningful snow event are much greater than without it. On the other hand, a positive AO and NAO combination typically supports a warmer pattern, with a storm track that will often go to our west.

It is difficult to predict far ahead of time, but we think the AO and NAO will average more toward their negative phases this winter. February and March present the best chances for a negative AO and NAO to lead to a colder/snowier pattern.

We’re leaning toward a negative AO and NAO, as the indexes have averaged negative since the spring, and we’ve seen somewhat persistent high pressure over Greenland.

Links to past outlooks and evaluations