Model Diagnostic Discussion
NWS Weather Prediction Center College Park MD
245 AM EDT Thu Aug 08 2019

Valid Aug 08/0000 UTC thru Aug 11/1200 UTC

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

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

...Overall Pattern Across the CONUS...
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Preference: General model blend, limited weight on the GFS by day
3.
Confidence: Average

07Z Update: The large scale features remain in very good agreement
amongst the various global models, with only smaller scale
differences creating discrepancy in the pattern evolution. The
most notable difference involves the magnitude of the mid-level
ridge across the central CONUS by day 3. While the GFS shows a
much more rapid breakdown due to shortwave energy rotating around
the periphery, it is alone among the deterministic guidance in
this evolution, as the others keep the ridge amplified through the
end of the forecast period. Although this has some ensemble
support from the GEFS mean, the mean is considerably weaker with
the de-amplification of the ridge, more in line with the remaining
guidance. Despite this, QPF differences are minimal, so for
continuity with the previous blend will not remove the GFS, but
suggest limited weight by day 3.

Previous Discussion:
Model mass fields are in very good agreement through the day on
Friday and into early Friday night, with the main issue being
smaller scale features which impact QPF placement and magnitude
along a low moving cold front over the south-central CONUS.
Similar difference appear once again ahead of the next cold front
over the northern Plains on Saturday. A broad-based model blend is
preferred for mass fields across North America...with more focus
on higher resolution guidance for QPF.

Differences in two parts of North America emerge Friday
night/early Saturday. One region is over central and eastern
Canada with implications for the magnitude and speed of
propagation of a mid- and upper-level trough over New England. The
NAM was a slow/strong outlier in moving the trough from New York
across New England on Saturday...while the GFS was clustered with
other operational and ensemble means. 

The second area of differences in the synoptic scale on Friday
night is the progression of a trough into the West Coast. In this
case, the 00Z GFS remained the most progressive...which has
implications in the shortwave trough progression downstream around
the northern high plains ridge axis to the trough over the
northeast CONUS. Am tempted to put less consideration in the GFS
idea due to convective feedback appearing in the mass fields on
Day 2...but the impacts in terms of QPF is minimal suggesting the
model blend can include the GFS without significant problems.

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

Bann/Weiss