Some Model Analysis
It was incredible to watch the computer models dance around this for days leading up to Wednesday and then suddenly go bonkers from DC North. Kudos to the NAM model for picking up that this was going to be a VERY juicy storm. It scores another coup this month. And one thing I've noticed, is that when it comes to some of the bigtime events in the East, this model tends to do VERY well. This occurred on a notable scale around Valentine's Day of 2007. It started upping the amounts of precipitation expected right away. Began at 2-2.5" and ran it up to 3-3.5" with time, and absolutely nailed the corridor that took in the most snow ultimately. The NAM also did a VERY important thing during this storm. It consistently showed that on the north and west fringe on the snow shield, there was going to be a very sharp cut off between little to no snow and warning criteria snowfall. This verified remarkably well. We can argue there were some issues with placement of it, but major kudos, as this is something the GFS model lost as we got close in. So while the NAM model is maligned in many respects, and rightfully so, it has proven during two of the most recent historic events in the Northeast to have a formidable amount of skill. The NAM will flop on a major snowstorm in the future...that's destined. But in the meantime, if it does show a unique solution that makes some sense, then I think one needs to take it into consideration when forecasting.
The Calls
Here is the order of my forecasts/calls issued chronologically. Click any image to enlarge.
Initial Concerns:
First Accumulation Call:
Second Call:
Third Call:
Fourth/Final Call:
Modeled Snow Totals through 7 AM EDT 12/20/09
All in all I'm rating my third call the best. Just some quick highlights...
Initial Call: Issued Wednesday night. Worked well in DC and VA. Nailed the corridor of max snow that occurred in DE and NJ. Nailed the mixing on the coast. Did not work in that I had the mixing too far north, left out LI, NY and New England. Did not hit the max that occurred in S PA and especially SE PA. Overall a D for missing NY and New England.
First Accumulation Call: Issued Thursday afternoon. Worked well again in DC/VA. Underdid amounts, but not terrible for an initial call. Successfully included NY and New England. Hit on max potential in NJ again. Once again nailed the mixing, but again was too far north with it. Underdid amounts across the board in the main area of snow. Was too far north with the 2-6" range. Overdid snow amounts on Delmarva. Overall I'll go C+ here, pluses for lead time and continuing to do decent in the Mid-Atlantic, but minuses for faltering a bit in New England/Long Island again.
Second Call: Issued Thursday Night. Backed off a bit on totals because of model trends. But essentially had a similar map to Call 1. A little bit more an expansive map. Overdid parts of Ohio here. Was way too bullish in PA/NW NJ north of I-76/78. Better in Delmarva. Left out the max total bullseye in NJ/DE. A downgrade from the first call. D.
Third Call: Issued Friday afternoon. Probably the best map of the event. From north to south. Did very well in Southeast Mass. deliniating the line between Blizzard conditions/totals and moderate totals. However, missed a good chunk of Northeast Mass. up into Maine. Wasn't quite far enough west in Central Massachusetts. Was decent in Southern RI, but not so much in N RI. Overall I needed to lift my bands about 45 miles further NW in Mass and the max bands about 20 miles further NW in RI/CT. Did very well in Western CT. Did well on the Western half of Long Isl, but grossly underdid the eastern half. Did well in NJ for the most part, except was too bullish with mixing to the North. Didn't do terribly in PA with the cutoffs, but needed to be a little more northwest in SE PA. Maybe a little too aggressive still on Delmarva, but did decent in MD/VA and WV. Overall, I'll give it a solid B.
Last Call: Should have just kept the initial call and modified New England. Not a great update. Give it a C.
All things considered...room for improvement, but I'm content that I didn't at least flop mega-time.
I outlined some specific totals in calls 3 and 4 (modified slightly from the maps). Let's see how those worked out... (The first range in white is the forecast #, the blue colored text represents the actual for the station, as well as Metro/Area totals near that forecast point):
Call 3
RIC: 6-10" (7.4"... 6-14" area-wide)
ROA: 15-25" (17.8"... 15-20" area-wide)
DCA: 12-20" (16.4"... 16-23" area-wide)
BWI: 10-18" (21.0"... 14-23" area-wide)
Baltimore: 10-18" (17"... 12-20" area-wide)
Wilmington: 10-18" (17.0", 13-18" area-wide)
Philly: 7-14" (23.2"... 14-24" area-wide)
ACY: 12-20" (12.1"... 10-18" area-wide)
MIV: 12-24" (16-24" area-wide)
Cape May-Seaside Heights: 7-14" (4-10" Cape May, 12-20" Ocean County)
Dover: 12-24" (18.0", 13-22" area-wide)
SBY: 5-10" (4-6" area-wide)
TTN: 7-14" (7-14" area-wide)
ABE: 3-7" (5.6"... 3-6" area-wide)
MDT: 4-9" (8.8"... 6-18" area-wide)
AVP: 1-4" (0.3"... 1-4" South and East)
NYC (Bronx/Yonkers): 6-12" (8-10" Bronx, 10-12" Yonkers)
NYC (Queens/JFK): 8-16" (10.9" CPK, 8.8" LGA, 14.2" JFK)
Hartford: 3-7" (3-7")
Providence: 4-9" (15.1"... 14-22" area-wide)
Boston: 3-7" (10.0"... 8-14" area-wide)
Suffolk Co, LI: 10-20" (14-28")
Nassau Co, LI: 8-16" (12-18")
Morristown: 4-8" (4-10")
Call 4
RIC: 7-14" (7.4"... 6-14" area-wide)
ROA: 20-30" (17.8"... 15-20" area-wide)
CHO: 18-28"(21.0", 18-24" area-wide)
DCA: 15-24" (16.4"... 16-23" area-wide)
BWI: 14-22" (21.0"... 14-23" area-wide)
ILG: 14-22" (17.0", 13-18" area-wide)
PHL: 10-18" (23.2"... 14-24" area-wide)
ACY: 13-23" (12.1"... 10-18" area-wide)
MIV: 15-24" (16-24" area-wide)
Beachfront Towns in: DE-Cape May Co NJ-Atlantic Co: 7-14" (4-6" Cape May to 12" Atlantic City)
Mainland Towns in above locations: 10-20" (10-18")
DOV: 15-24" (18.0", 13-22" area-wide)
SBY: 5-10" (4-6" area-wide)
TTN: 10-20" (7-14" area-wide)
ABE: 5-10" (5.6", 3-6" area-wide)
AVP: 2-5" (0.3"... 1-4" South and East)
MDT: 5-10" (Higher amts in Lancaster/York) (8.8" MDT, 12-20" Lancaster/York)
NYC//LI: 7-14" N & W, 10-20" S & E (8-16" N & W, 14-28" S & E)
HFD: 5-10" (3-7")
PVD: 6-13" (15.1"... 14-22" area-wide)
BOS: 5-10" (10.0"... 8-14" area-wide)
MMU: 6-12" (4-10")
Overall, a good learning experience for sure. Pleased with how things worked out for the most part. Thanks to everyone who played along and gave me some snow totals and basically encouraged me to keep outputting maps. Here's to the next one happening in late February when I have nothing planned to do with myself!
