In his D.C. area forecast through the weekend, CWG’s Matt Rogers noted that his forecast for Sunday was a low-confidence one. Looking further ahead, that same low confidence applies all the way through the Dec. 17-24 period, as the overall pattern appears to be in transition.

Temperatures during the period are expected to be closer to the seasonal norm than we’ve experienced so far this month, with below-average temperatures expected toward the days leading up to Christmas. Highs in and around the city during the Dec. 17-24 period typically average in the mid-40s with lows in the low 30s (lows out toward Dulles Airport typically run 5 to 7 degrees colder). But the area may see a couple of chillier days somewhere in the Dec 20-24 time range with highs around the upper 30s to low 40s and lows around the mid-to-upper 20s.

Could the colder pattern mean snow? The models are suggesting the possibility of a storm sometime around Dec. 18-19, but differ on the timing and track of the storm with some runs taking it well to our north and others to our south, sometimes far enough south to miss us completely. Model projections that far in the future are not very skillful, of course, but this potential storm is worth monitoring as the forecast pattern is not as bad for snow lovers as it has been

Technical discussion for December 17-19

Both the European and GFS model mean patterns for the 6-10 day period are now indicating that the higher heights and blocking that has been over Iceland and eastern Greenland, a little too far east to suppress the jet stream and send cold air down the East Coast, has finally shifted west enough to start lowering heights across our area. The block (red area) over eastern Canada may finally force an area of below-normal heights (blue) across portions of the South. Such a forecast promises to bring us cooler weather than we’ve seen this December so far.

Despite the similarities in the forecasts depicted below, there are also subtle differences. For example, the GFS mean brings the lower heights farther south and east along the Southeast coast than the last night’s European model. Those differences lead to variations in how the models handle a storm during the period.

5-day mean 500mb height and anomaly pattern ending 7PM Wed. Dec 19 from the European model (left) and GFS model (right). Red areas indicate above-normal heights and blue areas indicate below-normal ones.

Last night’s European model is slower bringing the storm eastward and tracks it farther to the north than either of last night’s GFS model runs. The European model track to our north keeps us in the warm air ahead of the storm longer than the GFS. However, today’s European model now tracks the low off the Southeast coast, more like how last night’s GFS model runs took the low to our south and gave us a cooler look for the storm. Today’s GFS keeps the same southern track and has trended even colder, but still keeps surface temperatures above freezing throughout the storm.

06Z GFS mean sea level pressure, 2 meter temperatures and precipitation forecast valid 1 PM Wed Dec. 19.

The various GEFS ensemble members have a wide range of solutions, some taking the storm north of D.C. and others to our south. The 06Z GFS (see right) which has a track that is usually favorable for snow this time of year keeps the boundary layer temperature well above freezing with the freezing line anchored in northern Pennsylvania during any precipitation that falls, suggesting that the storm is more likely to be a rainstorm than a snowstorm if it even occurs (that’s still a huge if).

Given the high uncertainty, pretty much all options are on the table for now

Technical discussion for December 19-24

The D+11 ensemble mean forecast centered on Dec. 22 suggests that once the Dec. 18-19 storm shifts to our east, temperatures will probably average below normal for several days with highs in the upper 30s to low 40s and lows in the mid-to-upper 20s. Composites of thee 10 analogs strongly favor a period when temperatures will average below normal for the period Dec 20 through Christmas Eve (see below). However, those same analogs are relatively dry, so below-normal precipitation is likely lowering our chances of getting a white Christmas. Only one analog date (Dec. 26, 2010, when a powerful Nor’easter largely bypassed D.C. but buried much of the Northeast with more than a foot of snow) yielded an inch of snow during the 5-day period

Temperature anomalies based on the CPC D+11 analog dates. The right hand side is for the dates of centered means from the analogs (equivalent to Dec. 22) and the left hand side is two days earlier.

The bottom line

The pattern is slowly transitioning to a cooler one after sometime around December 19 or 20. The models are suggesting the potential for a storm to pass close enough to us to produce precipitation sometime next week prior to any significant cool down. For now, there’s too much uncertainty to favor rain over snow or vice-versa, that is if the storm even occurs in the first place. That we’re even talking the possibility of snow, though, is a change from earlier discussions, so stay tuned.