Capital Weather Gang’s forecast for the winter storm Wednesday night and Thursday – which we named Snochi (based on reader suggestions) – was quite good, but not perfect.

At the airports, the total snowfall was 7 inches at Reagan National, 11.5 inches at BWI, and 13.3 inches at Dulles. Generally, in the D.C. and Baltimore regions, about 6-12 inches fell along and east of the I-95 corridor, and 10-18 inches fell west of the I-95 corridor (with some isolated amounts of 18-21 inches in Loudoun and Frederick counties).  In southern Maryland 3-6 inches fell.

Snowfall totals according to the National Weather Service

Link: List of snowfall totals

When you compare these actual totals to what we forecast (below), you see we were spot on in southern Maryland and east of I-95, but too low west of I-95 by up to several inches.

Capital Weather Gang snowfall forecast issued 2 p.m. Wednesday, 2/12.

Many areas right along and west of I-95 reached our “boom” threshold – which we said had a 20 percent chance of happening.

Generally speaking, we did a nice job presenting the overall pattern of snow and where the heaviest would fall. We correctly indicated there was a slightly better chance of a boom than a bust.

Also, I think we did a good job telling a steady story – which was more or less right in the 36 hours leading up to the storm – and not wavering.  Our first snowfall map issued Tuesday at noon (below) was very similar to our final forecast (above).

Capital Weather Gang snowfall forecast, first issued noon, Tuesday, 2/11.

The adjustments we made to the final forecast (two maps above) – to increase the area likely to get 8-14 inches and decrease the bust probabilities – were in the right direction.

Water vapor image of storm as comma head passes over D.C. (Stu Ostro from CIMSS, University of Wisconsin) Water vapor image of storm as comma head passes over D.C. (Stu Ostro from CIMSS, University of Wisconsin)

Other things we did well

* We did a good job indicating when the snow would start. We consistently said it would start in the 7-10 p.m. timeframe, after Wednesday’s PM rush hour, and it did exactly that.

* We correctly noted snow would fall very heavily Wednesday night and that there would be a dry slot mid-to-late morning Thursday into the afternoon, with precipitation tapering off – especially in the immediate area and points east.

* We knew the forecast was tricky for the storm’s second wave late Thursday afternoon and evening – it was tough to pin down if and where a final band of heavy snow would come through.  I think we did a nice job communicating that – including a detailed technical discussion from our winter weather expert Wes Junker.  Ultimately, we called for a closing round of mixed precipitation and then snow starting between roughly 3-4 p.m. and ending between 9 and 11 p.m., with 1-4 inches of new accumulation.  That was the right call although you could quibble that we waffled a bit as to who was most likely to get closer to 1 inch versus 2-4 inches (we first said more west, then shifted to more east).

* On Sunday afternoon, more than 4 days before the storm, we noted the European and Canadian models showed potential for the storm, while the GFS model did not. We  wrote then: “We slightly favor the European stormier scenario (as it has been consistent) but – even with that – pinning down specifics is difficult as the region will probably be close to the line between more of a snow event or more of a mixed precipitation event (snow to sleet to rain). As usual, colder north and west suburbs will probably have the best chance of more snow.”  Not a bad assessment at that point.

Where and why we erred

* We under-forecast snowfall totals along and west of I-95 mostly because we did not totally buy into the very aggressive simulations from the NAM model (and some others) – which turned out to be just about right. We probably compromised on totals a little too much by taking into account the (wrong) GFS model which showed considerably less snow.

* The snow changed over or at least started to mix with sleet and rain somewhat sooner than we expected along the I-95 corridor and points east.  Also, the mixed precipitation reached all the way into parts of Loudoun county, a bit farther west than we thought.  The earlier and farther west mixing occurred because the storm tracked slightly farther west than we were thinking, drawing in more warm air off the ocean.  But because the snow fell so heavily before the changeover, this little shift didn’t impact forecast snowfall totals meaningfully.


In what was a complicated forecast, I think we did pretty well.  I’d give us a B+ — so a silver medal effort.  Maybe we can earn a gold on the next event?

We welcome any feedback on our forecast or our presentation of forecast information.