The Winter Forecast for Delmarva, Philly and Baltimore/DC

By WBOC Chief Meteorologist Dan Satterfield

So, here it is! My take on the winter ahead, but first some words of warning.

  1. I’m probably wrong! Long range forecasts have low skill and are often wrong.
  2. A specific day by day forecast beyond ten days is not possible. This includes the forecasts in the Farmer’s Almanac and the 45-day forecasts on the AccuWx App. They are bunk and are as reliable as a horoscope or getting your palm read!

The scientific technique for making a long-range forecast is based on some real science, and chief among these is the scientific fact that the oceans hold almost all of the heat!

In short, the oceans run the atmosphere!

If we look at ocean patterns and then find similar patterns in the past and look at how the winter turned out, we can often have some skill in predicting the temperature patterns of the coming winter. Often, these patterns evolve during the winter, and this can lead to some large errors, but it is the best we can do. The oceans and the atmosphere always closely coupled and the biggest ocean patterns like El Nino and La Nina are feedbacks between the ocean and the atmosphere.

It’s Going to be a La Nina Winter

ENSO_LaNinaWintersSince1950_temp_maps_lrgAny forecast for the winter ahead should start with the state of the El Nino/Southern Oscillation (ENSO). This ocean pattern has a big impact on the storm track and we now have a weak La Nina developing and it looks like it will last through the winter. This is really not good news since the La Nina signal for weather is not as strong as the La Nina signal. In addition, look at the chart below and you see that there is a rather notable difference between weak and strong La Nina events, so looking at an average of all La Nina winters can lead you in the wrong direction!

Weak La Nina events like this winter often trend near normal to cooler than normal, but there is a great deal of variance. So this is not a lot of help unless we can find something else. One atmospheric pattern that gives us a very strong signal is the North Atlantic oscillation the NAO. This pressure pattern in the Atlantic has a positive phase and a negative phase. A negative NAO happens when a strong high pressure develops over Greenland and this “block” in the flow causes cold air from the Arctic to be steered into the eastern third of North America. Unfortunately, it is difficult to predict the NAO more than 15 days out but there is one thing we can look at: the QBO!

The QBO is the Quasi-Biennial Oscillation and despite the crazy name, it’s easy to understand. The winds high in the stratosphere (around 15-20 miles high) tend to blow from the west for a few years then turn eastward. This happens on average at around two-year intervals and while last year we had a westerly QBO it looks like it is rapidly turning into an easterly or negative QBO phase.

So, what if we look at weak La Nina winters that had an easterly QBO? We can find those years and then average out the temperatures over the winter, and when you do, you get the graphic below!


Some of the winters with a weak La Nina and an easterly QBO were rather famous for cold air. Like 1981-82-1983. The winter of 2014 was pretty snowy here on Delmarva as well.The winters in weak La Nina years are rather dry though as this chart from the NWS in Washington shows. Moderate La Nina’s are quite a bit wetter. Keep in mind though that this is not looking at weak la Nina winters with an easterly QBO.


CaptureThere is also research that was published in Feb. 2016 that shows a connection between the QBO and the NAO. An easterly QBO can favor a negative NAO which, remember, is a strong cold signal for our region in winter.

There is also some controversial research that is getting more mainstream, and this is how the loss of Arctic sea ice is perhaps impacting the storm track in winter. Research indicates that the ice loss is causing the NAO pattern to trend more negative with colder air in the eastern U.S and parts of Europe. Other research looks at October snowfall in parts of Asia but this is still uncertain, and I have increasing doubts about its usefulness, so I am not including it in my forecast.

The Elephant in the Room

Look at the chart below:

Climate change cannot be ignored in a long-range forecast. It is indeed the elephant in the room, as temperatures globally are nearing 2° F warmer than 75 years ago. The smart call is to modify any forecasts I come up with to at least a bit warmer, but I must also factor in that the loss of Arctic ice may be impacting the NAO and causing strong Arctic air outbreaks in the northeast U.S. Climate change plays another role: the warm oceans. If a bitterly cold Arctic air mass moves over the region and a coastal low develops, that storm could have a lot of warm wet air to make snow with. Ocean temps. off Delmarva are indeed very warm right now. If they stay that way all winter, the potential for a historic snow event would be increased IF the cold comes in.

So, with all of the above, I think we can say that this winter will be colder and snowier than last winter, but all in all an average winter is most likely. If it gets really cold though, watch out for a real snow event that might go into the record books. I also think late December and January will be the snowiest and coldest based on the past winters with weak La Nina and easterly QBO events. I’ll forecast a 50% chance of normal snowfall with a 30% chance of above normal and a 20% chance of below normal snowfall. As for temperatures, I’ll forecast a 40% chance of normal temperatures with a 25% of below normal and a 35% chance of above normal for the winter months of December through February.

Oh, and remember, long-range forecasts are often wrong!