Model Diagnostic Discussion
NWS Weather Prediction Center College Park MD
152 PM EST Wed Dec 19 2018

Valid Dec 19/1200 UTC thru Dec 23/0000 UTC

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

12Z Model Evaluation...with Preferences and Confidence
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

...Overall Pattern Across the CONUS...
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Preference: 12Z ECMWF, 12Z GFS, greater weight to ECMWF
Confidence: Slightly below average

---18Z UPDATE---
No change to the preliminary preference. A blend of the ECMWF and
GFS still seems to offer a consensus forecast supported by
ensemble members.

---PREVIOUS DISCUSSION---
The largest area of precipitation over the next few days will be
generated by a broad and strong cyclone that will develop in the
Southeast and then move inland up the East Coast. Models generally
agree on the large-scale evolution, but detail variations make a
substantial difference with QPF intensity and placement, as well
as precipitation type on the cold side of the cyclone.

Model preference was weighted most strongly toward the 00Z ECMWF,
which has shown very good consistency with previous model cycles
and depiction of weather impacts. The 12Z GFS has generally
trended in the direction of the ECMWF, continuing a multi-run
trend, and these two models should form the basis of a reasonable
blend that is supported by a considerable number of ensemble
members.

The CMC and NAM were not preferred as much, given some slight
differences that do make an impact. The 00Z CMC ejects the
negatively tilted mid-level wave up the East Coast faster than the
other models leading to changes in the trajectory of the surface
low and placement of QPF. Given the anomalously strong nature of
the digging wave, and strong cyclogenesis expected, the preference
is toward the model consensus and a slower ejection of the wave
given that more closely matches the scenario. The 12Z NAM,
meanwhile, develops the surface low much further southeast
relative to the other models, kicks it northeast into the
Carolinas faster, and eventually cycles it much further northwest
toward Lake Huron than the other models. The preference is to
again lean closer to model consensus, with an occluding surface
low initially forming closer to the strongest height falls
(northwest of the NAM surface low position).

For the East Coast cyclone, the 00Z UKMET also offers a reasonable
forecast that is generally similar to the ECMWF and GFS. However,
over the Rockies and Plains, the UKMET is strongest with secondary
troughs relative to other models. These troughs will be pushing
into an amplified ridge (upstream of the large East Coast trough),
and thus with expected destructive interference of these waves,
the expectation would be for a flatter flow pattern overall.
Therefore, the preference is still to lean closer to the ECMWF and
GFS overall.

On the West Coast, the next wave approaching the Pacific Northwest
will generally have QPF impacts in the medium range, just after
the scope of this discussion. However, The ECMWF seems more
reasonable with the initial effects of the associated surface low.
The GFS is much further north, while the GEFS Mean and ECMWF
Ensemble Mean are south, closer to the deterministic ECMWF.


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

Lamers