Model Diagnostic Discussion
NWS Weather Prediction Center College Park MD
1120 PM EST Mon Dec 24 2018

Valid Dec 25/0000 UTC thru Dec 28/1200 UTC

...See NOUS42 KWNO (ADMNFD) for the status of the upper air
ingest...

00z Model Evaluation...with Final Preferences and Confidence
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Preference: 00z GFS/18z FV3 and 12z ECMWF/CMC blend
    Weighting heavier to 12z ECMWF/18z FV3
Confidence: Slightly above average


Ensemble spaghetti analysis shows the ECMWF/ECENS members have
been quite consistent with the placement/strength of the digging
trof into the Southwest, ejecting into the Plains producing a
strong winter storm/deformation zone as well as a strong squall
line/cold front with heavy rainfall across the Southern Plains
into the Lower Mississippi River valley through midweek.  So given
this consistency, the 12z ECMWF remains a strong preference in any
blend, though still may have some typical slow/over amplifying
negative biases manifest, especially in the lingering of the
deformation/snow band in the Dakota/northern Midwest into late
Thurs/Fri.  Additionally, the placement of the wave shearing into
confluent flow across the Great Lakes in advance of the digging
Arctic trof, also seems sensible and fairly stable.   However, the
ECMWF is also a bit weaker with the kicker wave/digging jet streak
upstream and perhaps is a shade too weak given the strong Pacific
flow regime, recent larger scale pattern over the last few weeks. 
As such will include other guidance to help temper this mild
upstream uncertainty. 

This solution has been recently bolstered by a few recent cycles
of the FV3-GFS and in particular the 18z run, providing much
stronger confidence and in comparison to the 12z ECMWF...the wave
is a bit but shows a very similar mass field evolution,shape and
time (as well in the QPF field).  The 12z CMC like the FV3-GFS is
a bit faster and shows a similar evolution, but unlike the ECMWF
or even the FV3 for that matter, is uncharacteristically fast as
well as strong with the upstream kicker wave, and by the end of
day 3 may be a bit too aggressive, so blended with the ECMWF may
help settle each other's potential weakness.  

The UKMET is compact ejecting into the Plains but also typical of
bias fast, which as usual seems counter intuitive, this leads to a
near phasing/absorption into the Arctic stream wave.  As for the
upstream shortwave digging into the Southwest by 12z Friday, the
UKMET like the CMC is a bit fast; but unlike the CMC pumps the
Pacific ridge and sharpens the digging trough into a broad but
near concentric closed low by 12z Friday and is not favored.  The
00z NAM, like the GEFS members and prior GFS runs are east with
the surface wave and while the convection/cold front
timing/placement and shape look reasonable, the northern portion
of the storm/deformation zone is too fast and not favored at this
time.  

Interestingly, the 00z operational GFS shows a fairly sizable
shift in solution from prior runs, which is typical as the wave is
now within the more dense RAOB observational network, and now
while slightly faster (typical bias) looks very similar to the
FV3-GFS/12z ECMWF solutions. As such, have no solid reason to
remove it from the blend.   As such WPC CONUS blend is a 00z
GFS/18z FV3/12z ECMWF/12z CMC blend weighted heavies to the more
consistent, 12z ECMWF/18z FV3-GFS, with slightly above average
confidence (especially for the lead wave, perhaps average for the
second,kicker wave).


Model trends at www.wpc.ncep.noaa.gov/html/model2.shtml
500 mb forecasts at www.wpc.ncep.noaa.gov/h5pref/h5pref.shtml

Gallina