Preview of NWS' New Version of Forecast
This preview is not operational and should not be used for support decisions.

Prognostic Meteorological Discussion (PMD)

Issued by National Weather Service (NWS)

FXUS10 KWNH 191918

Model Diagnostic Discussion
NWS Weather Prediction Center College Park MD
317 PM EDT Sun May 19 2019

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

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

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

...Deep closed low crossing the Southwest through Monday....
...Going negative tilt across the High Plains by Tuesday...
...Leeside cyclogenesis over the High Plains...
Preference: General model blend...through 22.00z
            Blend of the 12Z GFS/ECMWF...after 22.00z
Confidence: Slightly above average

19z update: Consensus is still fairly good through 36-48 hours
across the Plains with the storm system expected. Beyond that,
timing issues develop with the NAM remaining the faster outlier,
and the UKMET on the slower/southeastern side of the guidance
envelope. A blend of the GFS/ECMWF seems most plausible at this
point, so will continue to prefer that for the WPC blend.

---previous discussion---
Anomalous closed low currently off the California coast will work
through the Four Corners region Monday then eject out into the
southern/central Plains Monday night into Tuesday, taking on a
strong negative tilt as it does so. It is then expected to slow as
it lifts north, phasing somewhat with another approaching
shortwave, forming a large gyre over the interior west. At the
surface, lee cyclogenesis will take place Monday, helping to surge
anomalous moisture northward. In the broad, large-scale sense, the
latest models are in rather good agreement with the setup and
evolution. The differences lie in the details, on boundary
placement, and surface low position, especially Monday evening
through Tuesday. The trend has been for a more westward position,
over the front range of southern Colorado. This favors the GFS,
but most of the deterministic models now have a similar position
through 36 hours. Beyond that, some take the low quicker northeast
than others, the CMC being the slowest while the ECMWF is on the
faster end of the model spread. But overall, it's not too
significantly different. As such, for the mass fields, a general
model blend can be used with some higher inclusion to the GFS and
ECMWF for periods after 22.00z.

...Vigorous Pacific closed low digging into the West Tuesday...
Preference: General model blend
Confidence: Average

19z update: Overall, no significant changes noted with the rest of
the 12z guidance now available. A general model blend still is

---previous discussion---
Another strong northern stream closed low will move into the
Pacific Northwest and California coast Monday and Tuesday. The
main PV anomaly will drop south through California, as it becomes
absorbed within the large scale troughing that will develop over
the western US. In general, the models show fairly similar
solutions through the next 3 days with its mass fields, exhibiting
some of their typical biases. The GFS is on the faster spread of
solutions, while the UKMET/ECMWF are on the slower end. The CMC
becomes very different/slow by Tuesday evening, and should be
discarded as an outlier. Overall, a general model blend should
suffice with this system over the next 3 days.

Model trends at
500 mb forecasts at