Model Diagnostic Discussion
NWS Weather Prediction Center College Park MD
411 AM EDT Tue Aug 13 2019

Valid Aug 13/0000 UTC thru Aug 16/1200 UTC

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

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

...Overall Pattern Across the CONUS...
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Preference: Blend of the 00Z GEFS mean/12Z ECENS mean
Confidence: Below Average

The general pattern evolution is of zonal flow becoming a western
ridge and an eastern trough over the next few days. This is well
agreed upon by most global guidance. However, the small scale
features remain quite uncertain, and this is continuing a lower
than average confidence forecast in the preferred blend.

The UKMET remains a strong outlier with its heights both within
the developing western ridge and the eastern trough (weaker trough
despite a stronger ridge). While the remaining guidance has
trended a bit more amplified with the 500 pattern, the UKMET is
still several dm out of tolerance with the global consensus, with
little in the way of ensemble support.

On the smaller scale, several shortwaves, some of which are likely
to be convectively reinforced and have little forecast skill on
large spatial scales at distant timranges, will rotate atop the
ridge through generally zonal flow before dropping into the trough
the middle of the week. The NAM remains strong with much of this
energy compared to the remaining guidance, while the CMC is again
several degrees of longitude too fast with these impulses, as well
as a more significant feature dropping into the PacNW Thursday,
despite considerable model spread.

This again leaves just the ECMWF and GFS, both of which are
experiencing their own set of issues with minimal consistency in
QPF/frontal placement on day 3. The means of the GEFS and ECENS
again appear to be the best solution at this time, both for
consistency, but to also remove some of the fine details that
struggle to be resolved in the forecast. While these small details
will be important, they are best reserved for a shorter-term
forecast (ie Day 1) and it is suggested the means be used
exclusively beyond 24 hours.


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

Weiss