For days, we’ve kept these model snow forecast maps close to the vest, as they are typically not reliable more than about 72 hours out, and sometimes less. But now, as we’re within the 72 hour window within which the snow should start Friday and we’ve told you double digit totals are forecast, we’re no longer hiding the goods.
Before the curtains open, we re-iterate the message that shifts in model forecasts are likely over the next three days and these amounts are certain to change. How much is difficult to know.
While there is remarkable model agreement that heavy snow is coming to the region, there’s still about a 1 in 3 chance shifts over the next few days could result in a moderate snowstorm or wintry mix mess rather than a mega-snowstorm.
So without further ado…
The GFS model (afternoon run): 17-35 inches
The GFS model ensemble mean* (afternoon run): 18-27 inches
The GFS model (morning run): 15-30 inches
The GFS model ensemble mean* (morning run): 18-27 inches
The European model: 5-30 inches
The European model ensemble mean*: 12-21 inches
The Canadian model: 15-22 inches
Note that these models assume 10 inches of snow will fall for every inch of liquid equivalent precipitation, which may be too much in areas near the rain-snow line where sleet and rain mixes with snow or not enough in colder areas where the snow will have a fluffier consistency.
The average snowfall from all of these simulations in the D.C. area is in the range of 15-20 inches.
But again, these model solutions will fluctuate some over the next two days. Totals as high as shown above may not materialize if either the storm track shifts substantially northwest, introducing more rain into the forecast, or southeast, which would result in less overall precipitation.
* Ensemble mean refers to the average result from a group of simulations from a parent model (i.e. the GFS and European), which has had its initial conditions and physics tweaked to better represent the range of uncertainty.