Forecast lows for Thursday morning from the GFS model. This model tends to run a bit colder than reality. Still, it’s only a matter of time until temperatures touch near freezing across the area. (

As the season’s first frost advisories Sunday night — and subsequent patchy frost Monday morning — probably reminded you, it’s that time of year again. Some spots across the western and northern suburbs of D.C. could dip to near freezing in the next several days. Even if they don’t quite make it this time, overnight temperatures near and below freezing are on their way to being standard fare.

Given the significant topographical differences in the D.C. area, and the proximity to large rivers, bays and oceans, that first freeze date has been known to come as soon as September to as late as December depending on location and year.

It’s a creeper-type of weather syndrome as well, moving from northwest to southeast as the season progresses.

Washington, D.C. and Baltimore metros

First freeze dates for the cold season beginning fall or winter of the year shown. We focus on the period since 1963 because we have records at each airport from that period onward. (Ian Livingston and Justin Grieser)
First freeze dates for the cold season beginning fall or winter of the year shown. The focus is on the period since 1963 because we have records at each airport from that period onward. (Ian Livingston and Justin Grieser)

Looking at the area’s most quality-controlled and consistent climate locations, the average first freeze makes its first train stop at Dulles, second at Baltimore (BWI), and third at Reagan National (DCA). Dulles’ first freeze averages Oct. 15, BWI’s usually comes along by Oct. 29, and DCA’s is about a month later than Dulles, coming around Nov. 18.

DCA tends to have its first freeze way later than most spots, heavily due to its location near sea level on a riverbank embedded in an urban heat island.

The graph below helps illustrate the often dramatic difference in first freeze date between Dulles and DCA during the 2000s. In 2001, for example, there was more than a two month difference! All this over a span of 23 miles.

Interestingly enough, the two locations have never had their first freeze on the same day since observations began at Dulles in 1963. The closest attempt came in 1971 when Dulles had a freeze on Nov. 4, and DCA picked one up on the 8th.

Dulles and BWI, on the other hand, have had had first freezes on the same day on several occasions. BWI and DCA have has well.

Mapping the broader multi-state region

Two years ago in a post on this subject, Justin Grieser, Katie Wheatley and I put together a map using data across the region from a much larger number of sources that are generally well maintained, but can have some continuity issues and are thus referenced less often than those observing stations located where nobody lives (airports).

That map is worth another look, to give a more visual sense of how the first freeze progresses across this part of the mid-Atlantic. You’ll note the significant impacts from high elevation (earlier freeze) as well as those from proximity to water and/or urban centers (later freeze).

162 locations were used to create this map which was subsequently further tweaked to try and present as accurate a picture as possible. While the impact of elevation is shown to some degree, we are likely missing some of that detail in the interpolation. (Map by Katie Wheatley. Analysis by Ian Livingston, Justin Grieser and Katie Wheatley.)

Since the map is an interpolation, with spottier data than we’d have liked in a perfect world where everyone keeps detailed daily temperature info for all of eternity, there are undoubtedly some errors in specificity. But… you get the idea.

Where we stack up on a central mid-Atlantic list

Yet another way to look at the variance across the broader region is to take a peek at a selected list of average first freeze dates, all of which went into the map above.

Selected first freeze average dates for 20 locations across the broader central mid-Atlantic. Some locations may have high variability over short distances, like near the Winchester, Va. area. (NOAA, adapted by author)

First freezes in the broader region more often range from early October through late November. The very highest elevations in West Virginia, not included in the list above, may more typically end up with a September freeze, while some coastal locations (and stubborn DCA) might even hold out until December.

One of the more interesting aspects of the list is how closely aligned D.C.’s “official” first freeze average date is to that observed in coastal southeast Virginia, with only Norfolk coming in at a later date.

DCA’s official first freeze is even two weeks later than Huntsville, Alabama’s!!

Of course, as city dwellers know, and as we’ve noted in prior first-freeze analyses, the District’s hills that rise to as high as about 400 feet – away from water and the most urbanized center – often see their first freeze up to several weeks before places like the National Mall or mushy bank of the Potomac River.

Given all the variance across our area, there are sure to be those who disagree with the patterns shown by the average first freeze maps, graphs and numbers above. Do these tend to jive with what you see locally? Anyone keep personal records?


When should the Washington, D.C. area expect to see its first freeze of the cold season? (2011)

First freeze of fall in the Washington, D.C. area is typical by mid-October (2012)

Reagan National Airport’s misleading low temperatures

Should Reagan National Airport remain Washington, D.C.’s official weather station?