REVIEW OF THE MODEL DATA 15 FEB 2012
2200 EST 15 FEB 2012
I am often asked a question like ” what does my gut say…?” or “what is my instinct about this particular event ?..” My response has been and will always be I do not forecast by gut or by emotion or by instinct . There are some other well known meteorologists — and a lot of weather weenies– who do that and it usually results in disaster. By committing to one particular forecast Model or idea days ahead of time … their bias becomes involved the forecast process. When one model comes out which shows what they think is going to happen… they Have TO latch onto to that idea it because they have decided days out to go with that particular solution. In other words by committing so early to only solution one loses the ability to think critically.
DT ‘S MEDIUM RULE #3… when ANY Model … which has been showing XYZ for many consecutive runs in a row… Suddenly shifts and in 1 run shows ABC…. The other pretty good that that shift is wrong… Unless all the other weather models also show the same sort of BIG shift.
The reason why this rule works is that in the current state of numerical model prediction — weather models– they are sufficiently accurate and of high enough quality that it almost all cases if one model shows a shift of significance the other models will fairly quickly follow suit.
This is why I NEVER Forecast by instinct or Gut. This is why REFUSE to even consider the GFS past 72 -84 hours hours when it comes to East Coast winter storms. Now that being said I am sure others e are going to manipulate or lie or change this assertion into something that I haven’t said. I do NOT have a problem with the GFS model for certain weather patterns OR once we are within 84 hours. The 0z 15FEB GFS model run was a massive …severe unprecedented shift… sSifting the coastal Low from the Georgia Coast … suddenly 300 Miles to the nw over ATL… from passing south and east of Cape Hatteras to passing west of Norfolk … fromn showing NO precipitation of any kind north a Washington, DC … to driving heavy snow and rain into New York City Albany and will Willaimsport PA. The fact that you may be living in the northeast big cities and the 0Z GFS is the picture perfect and one that you want to see for your location does it mean that the solution is a viable.
The proof ? all the OTHER weather models did not show any significant change in their track or forecast and the updated 6z GFS has gone back to the southeast U.S. coastal solution.
This is supported by the Canadian and the British models 0Z FEB 15 runs as well as the 0z ECMWF and the 0z European Ensemble…. All showed ABOUT the same SORT of thing” the surface Low tracking along the SE US coast in a ENE direction then over or close to Cape Hatteras. There was absolutely no supported at all for the ridiculous 0z GFS . None.
In fact the 0z FEB 15 European model had no precipitation of any kind north of Fredericksburg… The Canadian and the British models do not have any significant snow north of the Washington, DC Baltimore area. Again I Hope realize that reporting and discussing the weather models does NOT mean that the point I am Making is MY forecast.. I am CERTAIN that this will event wiill become a mjaor Precip event .
We saw the same sort of thing with the 12z GFS .. Another solution which was unsupported by even the GFS Ensemble . It became so bad that at midday 15 FEB HPC had issued a special statement regarding the bizarre scenarios and solutions being proposed by the 12Z 15FEB operational GFS run.
From HPC Model discussion-THE GFS APPEARS TO HAVE WORSE INITIALIZATION OF THE TROUGH’S PLACEMENT OVER BRITISH COLUMBIA AND NUNAVUT THIS MORNING COMPARED TO THE NAM/ECMWF. MEANWHILE…THE GFS IS MOST AGGRESSIVE IN ALLOWING AN UPSTREAM PERTURBATION CURRENTLY OVER THE PACIFIC TO CUT THROUGH THE SHORTWAVE RIDGE OVER THE CANADIAN ROCKIES ON THURSDAY…WHICH THEN PHASES WITH THE NORTHERN STREAM AND SLOWS THE ENTIRE GFS SOLUTION. THE COMBINATION OF WORSE INITIALIZATION AND QUESTIONABLE AND EVEN SUSPICIOUS PHASING OF THE SEPARATE STREAMS IS SUFFICIENT TO DISCARD THE GFS
But the Model fiasco last night with the 0z 15 FEB GFS and 12z 15 FEB GFS does emphasize again why I think the GFS model … with espect the East Coast winter storms … is a piece of crap.. Note that so far NONE of the other medium range large scale models has shown any sort of HUGE wild swings in 6 or 12 hrs time increments. NONE.
Increasingly this looks to me more and more as a Shenandoah Valley / western VA….and Northern Virginia snowstorm then up into DCA/ BWI area and possibly into southern NJ . Richmond I think is going to be more wet than white but I am not absolutely convinced of that all of the event is going to be all rain for central VA.
The area from Lexington south to Roanoke then down to Hillsville poses of what I think to be the biggest challenges forecast. Clearly areas which are in elevated terrain have much better chance of seeing or staying all snow and clearly areas which are and north of interstate 81 have a even better chance of seeing snow… Either in its entirety or a mixture of rain / snwo going over to snow as the cold air is pulled into the system. Once you move north of Lexington and Lynchburg there should be and of cold air east of interstate 81 so that the rain
Of course if one follows the 12z FEB 15 European model… And if you have access to the precipitation panels you will have seen that very little precipitation falls north of a line front Richmond to Roanoke Virginia. However the European Ensemble is much weather and has a large significant precipitation shield …with precipitation amounts of 0.25″ that extends to the Virginia West Virginia border as well as the Maryland Pennsylvania border and into Southern New Jersey .
There could of course the snow to the north of that… into Philadelphia and New York City. In fact I suspect is probably going to be even though as a deceiving most of the TV stations in those markets are not forecasting snow into those cities. That being said a honest interpretation of the pattern and what the model data has been showing in trending – again ignoring all 4 runs of 15 FEB GFS – 0z 6z 12z ans 18z – shows that this is not going to be tracking up towards Cape COD with a benchmark. That sort of thinking and analysis is just outright delusioal and shows a poor grasp of basic meteorology .
That is not to say there aren’t meteorologists out there who are forecasting that but the data does not support a track of the Low to the benchmark of 40N/ 70 W. And more importantly the overall 500 MB clearly does not support that a specially with a large closed 500 Low over se Canada.
































