Model Diagnostic Discussion
NWS Weather Prediction Center College Park MD
213 AM EST Wed Dec 12 2018

Valid Dec 12/0000 UTC thru Dec 15/1200 UTC

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

00Z Model Evaluation...with Preferences and Confidence
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

...Overall Pattern Across the CONUS...

Preference: 00z ECMWF/ECENS, 12z UKMET, 00z NAM
Confidence: Slightly below average

07z update: The 00z ECMWF still shows some significant differences
compared to the 00z GFS with respect to the closed low over the
southern Plains and southeast US later this week and weekend. It
continues to be the slower of the two models and in general, the
spread between the two has not changed much in the past few model
cycles. The 00z CMC did improve somewhat compared to the its
outlier 12z run, but still probably should not be included. As a
result, preferences will lean on the 00z ECMWF and its ECENS mean
along with the 00z NAM.

---previous discussion---
There are two focus areas across the CONUS where model variability
is higher. The first is with the digging wave that eventually
closes off across the southern Plains. Models diverge quite a bit
after 14.06z as the shortwave becomes closed over Texas. First
off, the 12z CMC is most obvious outlier as it is 1) too fast and
2) too far north as it progresses the low eastward. The 00z GFS
continued its previous run(s) being the faster solution in the
rest of the deterministic guidance, though it is trending slower,
becoming more in line with the NAM/UKMET/ECWMF. It is worth noting
the 00z GFS jogs the closed low toward the Gulf Coast while the
ECMWF/NAM/UKMET progress it more due east. The GEFS mean is more
similar to the deterministic ECMWF/NAM. The WPC preference here is
for a 12z ECMWF, 12z UKMET, 00z NAM blend. Some inclusion of the
18z GEFS is possible as well.

The other area experiencing temporal and spatial variability lies
across the Pacific Northwest with a shortwave pushing onshore by
Friday into Saturday. The GFS is again too fast as it lifts the
shortwave into Washington State and B.C. Friday night. Meanwhile,
the ECMWF is probably too slow as it's the only deterministic
model which closes off at 500 mb. The NAM/UKMET are more similar
at 500 mb. At the surface, there is even more variability with
what should be a very strong/deep low developing near 40N/-140W.
The GFS/NAM take this feature into B.C. very quickly while the
UKMET/ECMWF show a somewhat similar path as the GFS/NAM, but
considerably slower. The ECENS mean is probably a decent
compromise of these solutions.

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

Taylor