Prognostic Meteorological Discussion
Issued by NWS

Home | Current Version | Previous Version | Text Only | Print | Product List | Glossary On
Versions: 1 2 3 4 5 6 7 8
FXUS10 KWNH 231836

Model Diagnostic Discussion
NWS Weather Prediction Center College Park MD
135 PM EST Tue Jan 23 2018

Valid Jan 23/1200 UTC thru Jan 27/0000 UTC

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

12Z Model Evaluation Including Preferences and Forecast Confidence

...Broad Occluded Cyclone and Associated Trough over the
Northeast, Progressing to near Labrador by Wednesday...
PREFERENCE: General Model Blend
CONFIDENCE: Above Average

---18Z Update---
No change to the preliminary preference. Models remain in good
agreement overall.

---Previous Discussion---
Models are in excellent agreement with the evolution of this
system over the next 24 hours as it moves out of the Northeast
region. Available operational models show tight clustering with
the low position and limited spread with many of the mass fields.
Therefore, a general model blend is preferred with above average

...Northern Plains Shortwave and Associated Secondary Cold Front
Moving Quickly East to the East Coast by Wednesday Night...
PREFERENCE: General Model Blend
CONFIDENCE: Slightly Above Average

---18Z Update---
No change to the preliminary preference. Models remain in good
agreement overall, and several of the new 12Z model runs (UKMET,
ECMWF) initialized closer to the GFS.

---Previous Discussion---
There are some initial differences across the Northern Plains, but
it appears to have very little impact on the model scenarios
through the next 24-36 hours as the trough progresses quickly to
the east. The 12Z GFS initialized with slightly lower heights
across much of the Northern Plains than other models, but it still
shows a very similar amplitude and timing of the wave by 25/06Z
(Wednesday Night). It does show slightly faster timing of the
colder air over the Mid Atlantic, but overall is reasonably close
to other models. Therefore, a general model blend is preferred
with this system.

...Progressive Trough Pushing onto the West Coast Wednesday, and
Deamplifying as it Pushes East to Great Lakes by Friday...
...Associated Surface Low near the Canadian Border and Trailing
Cold Front Pushing into the Central US...
            Lean to ECMWF (65%) vs GFS (35%)

---18Z Update---
No change to the preliminary preference. The 12Z UKMET and 12Z CMC
have not trended enough from their 00Z runs to be included in the
preference at this time. The UKMET in particular still shows a
stronger surface low that is much further south of the other
models and ensemble means over the north-central US.

---Previous Discussion---
The 12Z GFS and 00Z ECMWF offer fairly similar scenarios through
the first three days of the forecast, with the most notable
differences emerging late in the period over the Plains. Despite a
similar low position by 27/00Z (Friday evening), the GFS has
faster timing of the cold front in the Plains, while the ECMWF is
slower, and in better agreement with most of the other models.
Even the GEFS mean (06Z run) is slower than the operational GFS.
Given the neutral or deamplifying trend of the trough through the
period, do not anticipate an accelerating cold front down the
Plains as we sometimes see in winter. Therefore, the blend
preference is weighted more toward the ECMWF, particularly at
later hours.

The 00Z CMC was not included as it suppresses the height falls
further west and offshore initially along the Pacific Northwest
coast, which leads to a flatter wave, a weaker surface low that is
situated on the southern end of model spread. The 00Z UKMET was
also not included, as it shows a stronger surface low (position
furthest south of any model), with noteworthy trough amplification
over the Plains -- unsupported by any other model. And although
the 12Z NAM is not significantly different with the timing or
position of the trough or surface low, it does have a cold bias in
the Northwest behind the cold front, which could lead to lower
forecast snow levels than would be suggested by other models.

Model Trends at
500MB Forecasts at


$$ is the U.S. government's official web portal to all federal, state and local government web resources and services.