We have just begun a stretch of three straight days in the 70s, in February. Flower buds are bursting and pollen levels are exploding.

It’s awfully tempting to say we are finished with winter. Some might even say it never began. But history argues not to count winter out during March, and weather models suggest that the ridiculously warm weather we’re witnessing now will not endure.

Winter may fight back.

March is a wild month

March is historically way too volatile to confidently proclaim we are done with outbreaks of cold air and snow before it has even started.

The month is known for powerful cold fronts and sharp temperature swings. In March 2014, Washington posted a high temperature of 69 degrees on the 12th but just 36 the next day.

It has actually been colder in March (averaged over the whole month) than February seven times (in 1887, 1890, 1891, 1909, 1932, 1960 and 1984) in records back to 1872. It’s rare, but possible.

Once in a great while, a blockbuster snowstorm socks the region. Remember the 1993 Superstorm that unloaded more than 14 inches at Dulles Airport? Or how about the 11.5 inches that fell in the city from March 28 to March 29 in 1942?

In three of the past four Marches, at least an inch of snow has fallen in Washington. And in the past 25 years, we’ve had at least four inches of snow in March six times.

The models are teasing cold weather for March’s second week

The weather of the next one to two weeks should be warmer than normal (the average high is near 50, and average low around 32), but model forecasts present a clear downward trend in temperatures.

The farther out in time, the less confidence we have in their predictions, but a return to chillier conditions is possible by March’s second week.

The group of simulations from the GFS (above) and European (below) models predict highs of only 45-50 starting around March 7, with lows in the 20s to low 30s.

Beyond the second week of March, forecasts are more uncertain. But Capital Weather Gang’s Matt Rogers, who specializes in long-range forecasting, doesn’t think the colder pattern will lock in. “It’s most likely transient,” he said.


So perhaps we’ll be in position to declare winter over around the middle of the month.

We’ve been burned declaring winter over too soon before

Because the weather in March and even early April is so volatile in the Mid-Atlantic, it’s dangerous declaring winter over too soon.

In 2014, two days after we announced that spring had arrived, some surprise snow fell March 30.

Last year, we declared winter over March 7, and we had a couple chilly days in early April (although our criteria for winter being over still held).

As a refresher, our criteria for pulling the plug on winter are:

*No identifiable threat of accumulating snow in long-range forecasts that could remain on the ground for 12 hours or more.

*No identifiable threat of a “cold weather” lasting more than 48 hours. We define cold weather as highs in the 40s (or colder) and lows below freezing in D.C. (as measured at Reagan National Airport).

A story at CNBC.com Wednesday led with a headline that said “meteorologists declare winter is over” this year.

But not this meteorologist. Not yet.