Model Diagnostic Discussion
NWS Weather Prediction Center College Park MD
1136 PM EST Sat Jan 19 2019

Valid Jan 20/0000 UTC thru Jan 23/1200 UTC

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

00z Model Evaluation with Preliminary Preferences
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

...Eastern U.S...
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Preference: Any of the operational models

A cyclone will track from the Mid-Atlantic states to offshore of
New England Sunday into Monday, driving a major winter storm over
interior locations and all of northern New England (including the
coast). Model solutions remain locked in with little spread on the
synoptic scale. The GFS continues to lean a little too progressive
in the upper air pattern across southern Canada, the Great Lakes,
and the Northeast U.S., but this does not make too much difference
until the end of Day 2 at which time precipitation will have ended
over New England.


...Western/Central U.S...
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Preference: Blend of the 00Z GFS with 12Z ECMWF/UKMET/Canadian

A broad but energetic trough of neutral to positive tilt will
enter the West Sunday and expand through the central U.S. to Great
Lakes region by Tuesday. There is some semblance of stream
separation, with distinct concentration of gradients over the
Northern Plains to Great Lakes and also from the southern Plains
into Mexico. Ensemble surface low clustering over multiple model
cycles has favored a low reaching lower Michigan by 23/12Z, and
the models in our preference are reasonably well clustered in that
vicinity. Regarding the depth and speed of the trailing southern
portion of the trough, the ECMWF Ensemble Mean has been trending
deeper with each run at the base of the trough over Mexico. The
GEFS is somewhat less deep and shows no real trend. Overall there
seems to be enough model evidence to suggest the ECMWF may be too
deep and slow. And confidence is lessened given the variability in
the operational EC the past couple of cycles. The 00Z NAM, on the
other hand, sets the more progressive edge of the envelope with a
faster surface in the lower Mississippi Valley. A lot of the model
differences are smaller scale and in the noise, but the NAM was
fast enough at the surface and flat enough aloft to be excluded
from our preference although it may be useful through Day 2.


...Pacific Northwest on Day 3...
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Preference: Blend of the 00Z GFS/NAM with 12Z UKMET/Canadian

Somewhat messy pattern here by Day 3 as multiple flattening
shortwave troughs attempt to get through a mean ridge position,
bringing some rain into the Pacific Northwest. Structure of the
shortwaves differs greatly among the models, but general
clustering of solutions as well as arguments based on advections
would suggest the ECMWF is too fast to move the northern-most
shortwave through the ridge and into the northern Rockies by
23/12Z. Therefore, we recommend a non-ECMWF blend.

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

Burke