2000  EST  27  FEB  2012….. STARDATE   201202.27

 

Exactly one week ago  — FEB 19– eastern Kentucky the southern half a West Virginia and Virginia experienced a significant snowstorm.  I have had a week now to think about   this event and to compile all of my forecasts… statements and present a thorough review.  Thus this  web page entry.

In addition I will explain why I argued so strongly that this was going to be a KY  WVA VA  snowstorm and that the system had little chance of turning up the coast into the big cities of the northeast.  As  we go  through this analysis  I will stress  several key points.

1.  The  pattern  strongly favored  this Low tracking ENE  off the NC coast and not coming  up the coast.   Unfortunately   many Meteorologists  and weather Hobbyists  along the  East coast  were   ignorant   of this and blindly  clung to the   old  saying
   ” the north trend “. 
2.  The GFS Model  performance sucked moose. The Model performance was appallingly  bad   in the period from  day 10/ FEB 10  to  Day2/FEB 17 with frequent  wild swings  from  solution to solution.  
3.  The close adherence of this seriously  flawed Model by   weather  hobbyist and professional meteorologist caused unnecessary Hedging and   uncertainty in the forecast.

 

The first statement was issued by me based upon the Day 10 12z European model February 10 run….  which showed a significant Low pressure area along the just off the southeast coast of North Carolina.  It turns out that the day 10 European model from FEB 10 was amazingly accurate and as close to a direct hit with respect to the actual surface   on February 19   for a  ten day forecast as  you have.  Of course the   FEB 10 European model   was off by one day but still  the  day 10 European  model  depiction  was pretty darn good.

FRIDAY  FEB 10  235 PM  I  posted    this:

* ALERT DAY 10 EUROPEAN MODEL HAS MAJOR EAST COAST SNOWSTORM … even for most of NC and all of MD and VA on  FEB 20

Of course several  times this winter we seen false alerts from weather models especially  at Day 8  Day 10  and Day 12.  However in this  case something was different with respect to the large scale or synoptic pattern.  The PRIMARY    reason WHY I believe this event had a better chance to become a reality  and a significant  event  … even at Day 10…  was the collapse  that was  forecasted by all the models of the super intense and large Polar  Vortex   (PV) over Eastern Canada.

I have said this many times before and I am sure I will say again.  The overwhelming amount of data shows that having a large powerful intense PV over eastern Canada— over Hudson’s Bay or  central or Northern Quebec / or Labrador–is an extremely negative or Hostile  synoptic  feature to development of major East Coast Winter storms.  Having the PV below 500 dm  in that geographical position is a  manifestation of the  + AO pattern.  That being  said … if  one  looks at the 500 mb maps  in  January and  early February you would  have seen a   large intense   PV  situated consistently over northern Hudson’s bay or Northern Quebec  or Baffin island.  Sometimes this feature has been as low as  474  dm and has been so large that it has  6  or  8     lines — height lines –around it.   That represents a lot of wind and  energy and becomes extremely hard to get the Polar Jet  (PJ)  to dive south and  merge or   “phase” with  the subtropical jet  (STJ)  over the Eastern CONUS.  And without the phasing…  you don’t get big winter East Coast storms.

In addition on the 10th I argued that  another key aspect to this  POSSIBLE event  happening was the cold front and significant Low pressure area tracking through the Great Lakes and February 17.  In order for this possible winter storm scenario to develop  the Great Lakes   Low had to move into southeastern Canada and become a  50 /50 Low.  This in turn would force the cold air south  into the Northeast   US and prevent the next system and tracking up into the Ohio Valley.

The    12z  FEB 10 european  had some partial support the next morning   from the 6z   GFS but  the model  truncation  after day 8/192 hour  clearly  caused the  flawed  GFS Model  to “lose ” the southern  system (the short  wave)  at the 500 MB.

The  12z  FEB 11  operational  European  model had the  southeast Low  but it was further south. The 12z  euro  ensemble  had the Low 200 miles further North  and this trend continued  with the 0z  FEB 12  run of the Operational European… and I posted   this  on the Facebook  (FB) page at 0155 hrs  12FEB

*** ALERT *** ABOUT FEB 19-20 THREAT ****  
The  NEW 0Z — early Sunday AM –   run of   the  ECMWF ( the european) shows a  CLASSIC VA -NC -MD SNOWSTORM  … but only  IF one looks at UPPER AIR maps .   But the surface map does NOT look all that good.   At 500mb  on day 8 – 7PM SAT FEB 18 the 0z euro has closed 500 Low centered over central ILL -IND…. the surface Low for  some reason is waaaay to the south which makes NO sense . The Model has the surface Low on the SC coast which is a tad  further North than the 12z Saturday run. BY the Morning of FEB 19 the closed 500 low is right over southern VA … and there is some snow falling  over central and eastern VA…. but again the Surface Low is waaaaaaay to the east which makes NO sense.

On the   other hand   the 0z  GFS   took the Low  into the Great Lakes  and  developed a   strong  SE  ridge.

However the 12z GFS  FEB 12  run  showed  a Major change or     “flip flop”  in that  the Model  took the Low  from MI  to the  GA coast and showed the   southern energy  — the short  wave–  weakening   rapidly as  it  propagated  east along  the  Gulf coast.

The 12z  Euro  run on 12 FEB  remained   very consistent. The model  detected   2   pieces of energy — 2  short waves -  over  TX and KS –  that   merged  or  Phased into  a  stronger  southern short wave  that developed a  Negative   tilt.  This in turn argued that the  surface Low  should be closer to the coast  than what the  Model  was depicting with its surface map for  FEB 19.

A SIGNIFICANT  development  occurred on the  0z  FEB 13 Model runs when the CMC or Canadian  model which had been showing no  development of the southern shortwave for the last several model runs…  turned it towards the European solution by developing a significant area of Low pressure along the Southeast US coast that bought significant snow into the southern half of Virginia.  Such a development of course began speculation about the new north trend as it almost always does but as we will see in this case that was not to be for number of different reasons.   On MONDAY    FEB 13    at 0121  hrs  I posted this …

‎0Z CMC MAPS FOR ALL TO SEE.. SIGNIFICANT SNOWFALL for southern VA standards NOT for say NYC or PHL or BOS standards into southern 25% of VA. This is BIG leap in what THIS Model was showing when compared to earlier runs… 

The trend continued with the 0z   13 FEB  run of the European model which showed it even bigger and more developed southeast US  coastal Low   tracking the low over or close to Cape  Hatteras. However the model began to experience problems with the northern and the precipitation shield.

Later that day  the  the operational run   of the 12z FEB  13 run of    GFS   Model came out with a dramatically different solution which caused a significant reaction with some of the more poorly informed meteorologists.  The GFS showed a major jump northward in its position of the southeast low so that it bought significant snow into the big cities of the northeast. Upon seeing this run I posted the following on the FB  page

* COMMENTS ON 12Z MONDAY 13 FEB  GFS run — REGARDING FEB 19-20 ** Ignore this run of the GFS ..it has huge error in it that Just forces the model to go off the reservation into idiot land.

This was followed by the assessment from  HPC/ NWS the declared the  12z 13 FEB  operational GFS to be seriously flawed bcause of corrupted data the getting into the model .

Of course that drew a very strong reaction from several sources both from whether hobbyists and from professional meteorologist who should have known better.  The statement from HPC was pretty strong and I don’t see how anybody could have possibly ignored their argument.

Leading the charge that the GFS solution was correct after all was of course  Joe Bastardi.  He made the argument that this  dropping  the operational GFS and its north trend was the same sort of mistake that  HPC made   with regard to the East Coast historic snowstorm   25-26  DEC  2010.  If you recall the  particulars  of that event … it brought 10 to 20 inches of snow from   eastern NC through  Norfolk across the lower Maryland eastern shore into the big cities of Philadelphia New Jersey and New York City and then into Southern New England.

Bastardi argument was of course absurd and had  little basis in reality but he counts on weather  hobbyists  not having a firm grasp of what the actual facts were in a particular situation.  The truth is that in the days leading up to the  25-26 DEC  2010 historic snowstorm   ALL of the models had   shown the same solution for several days.  This was not a case of one model showing ABC  while another model showed  XYZ   and another  showed  HIJ.  But more importantly on the day before the actual   25 DEC 2010   Major  East Coast Snowstorm …   all the model data   — the European …  the Canadian ….the British  AND  the GFS  ….and ALL of  the short range models for that matter as well — all took the potential coastal snowstorm   way out to sea .  It was  NOT  the case of a   ONE  model coming up with a bizarre solution and being dropped by   HPC.

In addition his analogy also fails because just as all the models both the global and shore range models  suddenly took the coastal storm out to sea on  the December 24 model  runs…  that Night  on Christmas Eve …ALL  of the models  suddenly  reversed course and bought and be major snowstorm for the immediate  coastal areas from Virginia to Boston.

Never let facts get in the way of good   Bullshit.

Fortunately even while the debate regarding the  obviously flawed  12z GFS  Model was raging on midday on the 13th of February   the 12z  Ukmet  came out and showed a  Major   se US Low tracking over Cape Hatteras and moving out to sea bringing a significant snowstorm to much of western NC …VA … and   southern Maryland.

1239PM  FEB 13  I  made this  post on  the FB page

**ALERT *** ALERT *** 12Z UKMET GOES BOOM.. shows major east coast snowstorm FEB 19…. for entire Middle Atlantic even VA and MD into NYC LI and southern  CT.

This was followed shortly by the  12z run  of the European model which developed the southeast low a little deeper and took it further to the north ride along the South Carolina North Carolina coastal areas with the track running from Charleston to Wilmington to Cape Hatteras.   This cause me to make the following post  on the FB page.

FEB 13  132  PM     ‎*** ALERT *** ALERT ***ALERT …. this is No drill … 12Z EURO brings  in MUCH stronger  SE US Low   with HEAVY SNOW to all of central and eastern VA –over 6 inchesand much of Northern NC …. POSSIBLE ACCUMULATING SNOW late on the 18th into the 19th… WOOF WOOF BARK WOOF…  

At this Point I tried to make the argument that  the overall pattern clearly did not support a north trend and that the system was not a threat for the big cities the northeast.  Despite the strongly worded statement  from HPC and    the  fact that  the   12z  European Canadian and British models all strongly disagreed with the flawed  12z 13 FEB  GFS   run…  there continue to be a lot of desperate hope   and  the strongly worded assertions that the  12z GFS  was  somehow … that HPC   was wrong ..  had to be at least take into consideration .

However there  was  another way to of solve this dilemma as to which model solution was valid and  which one was not.  That would be to look at the historical analogs and finding similar events as well as using your brain to understand the potential scenarios that a forecast of may consider.

There are specific reasons  or    mechanisms   in the atmosphere  that  determine WHY   Low pressure areas in the winter months that  form in the Gulf     will    turn up the coast  and why some do Lows  are  forced to turn out to sea near or to the south of Cape Hatteras.  This is a science and it is not supposed to  be based on whimsical  Nonsense.    These 3   particular instances or analogs of significant snowstorms   for the lower Middle Atlantic region highlight the reasons and the mechanisms as to WHY Low pressure   along the southeast   US coast might NOT come up the coast.

However if you are unfamiliar with these events and you do not understand the significance of the players on the field especially at   500 MB then you would mis-read the Models and end at making a bad forecast period

  3  CLASSIC  LOWER MIDDLE  ATLANTIC — NC    VA  southern MD   –  SNOWSTORMS OF THE PAST.

case #1  30 JAN 1995…

This first event from 30 JAN 1995… is a very close match to the event of February 19, 2012.  This first image shows the upper air map on the 30 for January, 1995.  As you can see   there is an ULL — Upper Level Low — over the Tennessee Valley in a manner that was very similar to the event of  19 FEB 2012.  But in addition there was a  strong   PV – polar vortex-  located over Eastern Canada.  The PV in the Middle of JAN 1995 was a very intense large PV…  in a manner of very similar to what we saw in the beginning of February  2012.

 

The SURFACE Map shows   the   se US Low tracking along with just off the Southeast U.S. Coast very close to Cape Hatteras.  These
precipitation maps from   29-30 JAN  1995… and    30-31  JAN   1995 showed that significant   snow  did fall over much of Eastern Kentucky the southern portions  of West Virginia and much of southwestern and Central Virginia.  Even more significant is that in this situation just like in  19 FEB 2012……   the   heaviest snows   fell over  sw VA  and the  southern  Piedmont.

 

Notice here that he northern jet stream does  NOT  drop or phase into the southern jet stream.  Keep in mind that it is the phasing of the two jet streams between the Mississippi River and the East Coast  which is what develops the coastal Low.   If the two streams do not phase there is no MECHANISM for the   southern short  wave or Low  to come up the coast.

 CASE  #2     2-3 JAN  2002… 

This is also a very close match to the situation  of   19 FEB 2012….  Because the winter of   2001-02 has been very similar in terms of
the overall temperatures and lack of  snow  over the eastern half of  the  CONUS….  And because the winter of   2001-02  was forecasted   to be a fairly cold and stormy winter  over the eastern  CONUS  just like  2011-12.

Here is the SURFACE MAP  from  2 JAN 2002 and it seems to be a near ideal setup for  significant snowstorm into the big cities of the northeast US.

As you can see from these upper air maps JAN  1-2-3 , 2002…  there is a large and very pronounced ULL that  was centered over Iowa and Missouri on January 1.  This feature dropped into the Deep South and then rotated off the Southeast U.S. Coast by January 3.  Note again that there   was  a very deep and very intense    PV — Polar Vortex –  at  480 dm or lower located over Quebec Canada.   Also … note the very strong Pacific trough crashing into the West Coast JAN 1-2, 2002…  which again bears a striking resemblance to the  500 MB     FEB 17-19, 2012.   The analog here is striking we close to what we saw last weekend over  TN KY  NV   and VA.

Notice here that  AGAIN the northern jet stream does NOT  drop or phase into the southern jet stream.  Keep in mind that it is the phasing of the two jet streams between the Mississippi River and the East Coast which is what develops the coastal Low.  If the two
streams do not phase there is no MECHANISM for the   southern short  wave or Low  to come up the coast.

Its Just that   simple.

And by the way …there was  widespread speculation and forecast hoopla that this   Low on the se  US Coast 2 JAN 2002  was also going to turn up the coast.     And finally here is the snowfall accumulation map. In this instance   the best snows were NOT over  sw VA and the
southern Piedmont but over central and southeastern Virginia.  However the principle and the upper maps remain quite relevant and analogous  to  19 FEB 2012.

CASE # 3…   25-27 MARCH   1971…

Again the surface maps appeared to be quite favorable for significant East  Coast snowstorm.  There  is a large arctic HIGH centered over the Great Lakes covering all of the eastern CONUS and a stationary front over Florida and the gulf coast into Southern Texas with significant precipitation developing on the front.  However by 17 March 1971 the significant surface low pressure area moves off the North Carolina Coast and out to sea by passing the big cities of the Northeast.    WHY did this Low  also  ”miss”   the big cities of the  NE and why    heavy snow ONLY  fall over   eastern KY southern  WVA   northern NC and   the southern half of Virginia?

 

The 500 MB maps clearly show what happened.  In this case we have a very nice   50/50 Low centered over Newfoundland  Canada and this feature is a strong signal for potentially East Coast snowstorm .  We also have a very pronounced   -NAO as well .   And the PV is not very strong and is displaced over North Central Canada.   On 25 MARCH 1971 –  the 1st map– we see a   Upper Low Located over   KS   and the  PJ –polar Jet – is  located along or just to the north of the was Canada border and drops into New England…  while the Southern jet stream can be found running from California to Texas and Oklahoma then into the southeastern states.

On 26 MARCH 1971…  the   weak disturbance in Kansas with the Upper Low has developed significantly and is now showing in negative tilt  in the short  wave trough   — SEE BLACK  LINE–  running from   eastern KY  to the Florida panhandle.  Notice here that   AGAIN the northern jet stream does not drop or phase into the southern jet stream.   By 27 MARCH 1971   the   southern   short wave or Low has moved off the Southeast U.S. Coast into the southwest portions of the north Atlantic.

And finally the snowfall map for this event in late March 1971 showed a large area of   8  to  14″ of snow  across much southwest …  Central … South central and into Eastern Virginia.   Less than 1 inch of snow   fell in  Washington, DC and  4″ of snow  fell in   Norfolk and Virginia Beach and 7″ fell   in Williamsburg.

==============================================

Significantly  after the    12z  13 FEB operational GFS  screw up… the  12z GFS ensemble on 13FEB was much closer towards the European Canadian and British models scenarios .  Indeed the GFS  ensemble  mean was also focusing on the rapid  propagation or eastward movement of an   next incoming trough into the West Coast and towards the Rockies.

The 0z   14 FEB  Global model  runs became a significant turning point in with respect to the probability of a significant  snowstorm  for   the lower Middle Atlantic region .  The   operational GFS showed a much more organized area of low pressure coming further north along the  southeast US coast and  had  snow/ rain up to the North Carolina Virginia State line for the first time.   This was followed by the  0z  run of CMC   ( the Canadian)  which showed a major snowstorm for    the southern half of  Virginia and southern  West Virginia.    The  0z European Model   run showed perhaps the strongest surface Low  yet  with  significant snows far north as Philadelphia  and Southern New Jersey  as the 0z European Model  has the Low  coming as far North  as  37 degrees N latitude before being turned out to sea.

However the 12z   14  FEB   European model shifted the Low  to the south again.  At this Point I began to focus more heavily as to whether not there  was  any chance for phasing to occur between the two jet streams.

All winter long the pattern has featured a very strong and progressive or fast flowing Pacific jet which has not really relaxed long enough to allow for any sort of serious cold air into reasons coming south from Canada.  Very fast progress of flows are resistant to phasing between the southern and the polar jet.  I stated as such on the  FB   page

FEB 14   144PM..
I have for the last few days of favored the flatter solution as a general overall outlook be cause of this particular winter pattern. The last two winters …if you farted too loud you get a snowstorm the East Coast because the pattern allowed phasing… in fact it forced a lot of phasing or merging between the two jet streams. This result in a lot of big East Coast winter storms.

 This winter as we all know has been a very mild winter because the mean pattern or flow has been coming in from the Pacific not from Canada and the arctic regions. Thus with  THIS   possible event … I suspect that constant Pacific Jet stream energy will keep coming in from the West Coast pushing the Southern System on an easterly or eastern northeasterly track and not allowing it to turn up the coast.

The critical issue is whether or not the strong piece of energy in the southern jet stream over Alabama Mississippi mergers or phases with the northern jet stream. If it does…IF… that would pull the system further north and Northern Virginia Maryland into fairly and New Jersey would see significant snow. IF it does NOT phase then this would favor a flatter solution increase the snow threat for central and sw VA into Northern NC

ON  14  FEB  I  Posted a NOTE on the FB page   “WHERE WE STAND  ON THIS POSSIBLE THREAT   FEB 19-20”.

Withe  respect  to the debate as to whether not the operational run  of the 12z  14FEB GFS    was   valid  or  BS…  recall  this run of the GFS  has the Low coming up the coast and dropping heavy snow from DCA to BOS …  was clearly proven by the  0z  15 FEB  operational GFS which did a another flip flop and took the  Low   back to the south… along the  Southeast U.S. Coast over Cape Hatteras and out to sea.

This new flip flop by the operational GFS   model cause me to make the following commentin disgust on the  FB page…

This is why I HATE the GFS past 72 hours when it comes to East Coast winter storms. The 0z GFS model run was a massive …severe unprecedented shift…   taking  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 … from   showing NO precipitation of any kind north  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… ALL OF THEM…. 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.

At this point in my statements to   my  clients …I again stated that for Central Virginia and the Richmond metro area this was probably going to be a rain event for at least a portion of the  event  but I could not categorically rule out the threat of snow if the trend for colder  temperatures continued.  And for southwest Virginia up through the Shenandoah this look like a pretty significant snowstorm .   One  of these statements was read  on air  during the morning news show on WRVA1140 AM by Jimmy Barrett.  This prompted one of the on air meteorologists  in the Richmond metro area   to comment that   ”despite  some rumors of snow you may have heard about for Sunday the precipitation will be all rain” 

The  0z  15 FEB    run of the GFS…  Amazingly enough had a another massive model flip flop and reversal…  and now took the  southeast coastal Low  well  inland of New York City and Philly which of course  forced the model  to produce widespread significant rain.   ON the FB page  I Posted the following

FEB 15   117AM  ….    ** ALERT **8 ALERT ** IGNORE THE 12Z GFS RUN … AGAIN ** its way inland andhas rain into NYC….  c onsider   the  GFS  last run 4 runs…   12z TUES flat Low moves off NC coast out to sea…  18z TUESDAY a tad further North… 0z WAY NORTH low tracks over ATL then west of Norfolk…  6z Flat wave off the NC coast …12z  way inland and north!.

After seeing the 12Z   runs of the European …British and ….Canadian models  I    finally committed to this publicly on the FB  page  at 126pm  15   FEB

GOOD SIZE SNOWSTORM LOOKING MORE and MORE LIKELY NORTHERN VA … NW VA Northern half of SHENANDOAH VALLEY MD DE   south NJ

This commitment   to forecasting a snowstorm for the lower Middle Atlantic    was rewarded by it more changes in some of the models late on the   15th of FEB… the 0z   16   FEB Model  runs.  As I have said  many …many times… the  operational GFS is a awful model when itcomes to forecasting East Coast winter storms.   It has a lot of trouble handling Southern  short waves…   it has a lot of trouble  Phasing    the  2  jet streams just to name a few.  But once you move within  84   to 72 hours of the event   the GFS was a much better model.

Sure enough  at exactly 84 hours before the start of the event on12z 19 FEB….  the  new   0z  15 FEB    GFS Model showed a dramatic turn towards the European solution.

at 1108pm FEB 15  i made the  following post  on the  FB page

‎**ALERT *** 0Z GFS COMING IN VERY DIFFERENT FROM THE 12Z GFS RUN and much more like the ECMWF / UKMET

This was supported by the Canadian model  0z  FEB 15   run  and like the  European and the new  GFS and  Ukmet   the Canadian model showed no significant snow north of the Washington, DC area.

The model trend continue to show the southeast U.S. Low moving over or just south of Cape Hatteras and out to sea and each one was coming in slightly colder than the previous run on all the models.  After posting the  12z FEB 16 GFS   model  forecast SNOW map on the FB page   I wrote the   following on the FB page   (  which should of  been  under  NEW NOTE     or  posted  here on the   web site )

16FEB 1205PM …   ** 12Z MODEL DISCUSSION 16 FEB MODEL…. VALID FOR FEB 19 EVENT***   and in this  thread I  stated  that so far No model  was bringing the Low north or   heavy snow into any of the big cities of the northeast.  In addition it looked to me that this was going to be a  sw   VA… eastern KY southern WVA    big snow…  And that the rain could turn to  snow for several hours in Richmond but the lack of cold air was going to be a problem east of the Blue Ridge mountains.

Shortly after I wrote that the new 12z  16  FEB  European model came in dramatically colder even in central and Eastern Virginia.  The European model snow map showed a large area of 40 inch snows even through the Richmond metro area.

131PM  FEB 16    ** ALERT *** ALERT ** 12Z EURO BRING IN COLDER AIR TO RIC CHANGES RAIN TO SIGNIFICANT SNOWFALL SUNDAY AFTER 1PM then moderate to heavy snow until late sunday evening … MODEL says 4 to 8 inches of snow Roanoke Charlottesville Lynchburg Richmond Fredericksburg BUT … BUT that is assuming a 10:1 ratio…. if this happens   the snow ratio will  wont be 10:1  around RIC

At this Point I decide to as time the issue a 1ST  GUESS    snowfall forecast  map   at   610pm 16 FEB .

As   we  moved  into the short rain model  timeframe    snow mounts on many the  models began to increase especially east of the Shenandoah Valley into the Piedmont and Central Virginia as the data turn colder and colder with each one.  This was especially true on the  0z      6z and 12z Model runs of the NAM   and  GFS.  It was  at this point that the  short  range models began to show significant snow developing Sunday afternoon across Central Virginia and continuing  to 1:00 AM on the morning of the 20th.

Significant rain events changing over snow with significant accumulations are not likely in Central Virginia but has been known to happen.  But more importantly  the reason for the colder temperature profiles on the shore range models was because the precipitation and the system coming out of the gulf was moving slower .  This in turn allowed for colder air to move southward lay on the 18th into the morning of the 19th.

On Friday morning February 17 I  did a live on air into view with  Jimmy Barret 1140am  WRVA.  My appearance  followed immediately after the  morning news broadcast which featured the some  clown from the  Weather Channel  named  Ray  Stagick.  I have no idea if this guy is a meteorologist are not   but  if he is…  is he  is a pretty awful one.  TWC  committed to inch of snow  or less at the at  the end of the event for the Richmond metro area.    As you can hear from the broadcast..I  went for  2-4  from the city    East and  4 to 8    west  and  sw  and nw of the city and  10″  in  sw  VA.   Still I suppose I should not complain about the weather channel too muchsince they make me look really good.

021712_Dave_Tolleris_1329486354_16656

During the  morning  hours on  15 FEB  all of the Local  TV stations  the their forecasted  snow amounts for Central Virginia in the Richmond metro area   to  1-2”  while  TV 12 continue to show less than 1 inch.  All the TV stations to agreed that western and southwestern Virginia would see several inches of snow and the media stations in the Charlottesville and southwest Virginia / Roanoke markets played up the threat of significant accumulating snow.  Unbelievably even as  NWSFO AKQ began to mention the possibility of accumulating snow TWC  –The Weather  Channel- still refuse to discuss any other outcome except for a rain event with a brief period as snow at the end across central Virginia.

FIRST  CALL MAP   was issued late in the day on  the 17^th.

It was based on the  short range data which showed a longer  delay with the arrival of the precipitation.  Most of the other forecasts and model data were   premised on the idea that the precipitation arriving   Saturday night before the cold air arrived.  However the shore range models continue to emphasize a delay with each passing run so that most the precipitation would not arrive until Sunday dawned over the Virginia North Carolina State line and rapidly expanding northward towards Maryland.

However this delay was significant because it  allowed the reinforcing shot of cold air to sweep into PA MD  WVA  and VA.  It was  not that the air mass was dramatically colder but the due points were significantly   drier.   The temperature profiles began to show that the precipitation might start as rain or as an Exe but a larger and larger portion of this event over Central Virginia was going to fall as snow  and NOT rain.  And over southwest Virginia the shore range models are really began to crank out some impressive snow totals.

Winter storm watches and warnings    were  issued by NWS  for all of southwest Virginia the southern third of West Virginia and Eastern Kentucky.

LAST CALL MAP   issued Saturday  18 FEB 7PM EST.

This was  where  I made my first serious mistake. During the last   2 runs of the  European model the model snow algorithm  showed significant snowfall  — 8-12”   in  west to  east  Band stretching from southwest Virginia/ Roanoke east across the southern Piedmont stopping just to the west of Richmond.   Even though most of the other short range models    on Saturday did not show Richmond getting over 8 inches of snow I decided to gamble as a war and extend the 8 to 12 inch band of snow into the southern Piedmont and fairly close to the western side of the Richmond metro area.

This was a mistake.  I thought that the model is increasing the  snow fall accumulations because   the model  has  detected  the colder drier air and the  later  start time which meant that more the precipitation  would actually fall was snow and not mix precipitation or rain changing to snow.

During the afternoon winter storm watch is were issued for all of Central Virginia for  3 to 6 inch snow amounts.

Lastly…  Surely after I issued the LAST CALL SNOW  FORECAST map the last row of the shore range models came out and a  dramatically reduce the northward extent of the precipitation and the amount of significant precipitation over central…  southwest   Virginia and Southern  Virginia.  This causes a reaction from  several uneducated weather hobbyists  –  called   weather weenies—that of course posted the radar trends did not support significant snow getting into central and southwestern Virginia .

AS  we  now know  the 0z   FEB 19  shore range model runs were all completely mistaken.  However the shift was so complete on all the shore range models that I decided to issue a reduced snowfall map.  Even if the 0z   FEB 19 shore range models  were wrong it was clear that was not going to  be a band of  8-12” snows from Roanoke extending into going through the southern Piedmont almost to the western side of the Richmond metro area.

SUMMARY

I believe that I quickly deduce the potential for this event earlier than anybody else based upon my knowledge and study of the  differences between East Coast snowstorms and lower Middle Atlantic snowstorms.

I recognize the progress  nature   of the pattern which would force the system to slide off the North Carolina Coast.

 My FIRST CALL MAP  was best and my forecast on WRVA 1140  was spot on.

 One other note…  I did we see several emails from folks complaining that the actual website had not been updated and appeared to be completely unaware that the updates…  were being posted directly at the FB page.  I will have to make efforts to ensure that annuity is aware of this feature.

ACTUAL    SNOWFALL…