Model Diagnostic Discussion NWS Weather Prediction Center College Park MD 247 AM EDT Sun Apr 07 2019 Valid Apr 07/0000 UTC thru Apr 10/1200 UTC ...See NOUS42 KWNO (ADMNFD) for the status of the upper air ingest... 00Z Model Evaluation with Preferences and Confidence ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ...Northern Tier of CONUS... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: General model blend Confidence: Slightly above average 07z update: The negatives of the 12z CMC/UKMET both seem to reduce with the 00z runs coming into better overall agreement. As such a general model blend can employed; confidence has increased to slightly above average. ---Prior Discussion--- Quasi-zonal flow, with multiple shortwaves associated with different airmass/frontal zones makes the mass fields quite muddled across the Great Lakes into New England. Yet, the guidance is fairly agreeable in the general evolution with the undercutting jet streak shunting the initial wave out toward SE Canada/Canadian Maritimes by late Mon/Tuesday. The old occluded low/shacked northern stream shortwave will then return across the Great Lakes on Tuesday and through New England early Wed. Overall, the CMC seems a bit more stretched out West to East with the slowest/lingering surface wave and associated shortwave upstream by the end of the forecast period. The 12z UKMET also shows a greater amplification by Day 3, having drawn shortwave energy back from the east across Quebec leading to southern and stronger closed low, building this deeper trof. Ensembles support the remaining guidance, weighted a bit more to the GFS over the ECMWF and lower to the 00z NAM. As such will support that blend but with so much small scale variability the confidence is only average. ...Southern stream system affecting Gulf Coast... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: 00z GFS/FV3-GFS/UKMET blend Confidence: Slightly below average 07z update: The 00z UKMET remained fairly consistent with the prior run. The 00z ECMWF did trend a bit deeper initially with the closed low but still shears the system faster allowing it to become uncharacteristically fastest shift the trof east. The 00z CMC slowed, retaining the strength of the closed low, hanging back with the NAM. While the ECMWF did trend favorably, it is still on the outer fringe of ensembles including the slightly faster 00z GEFS mean. As such will keep initial preference but remove the CMC, and weighting the UKMET a bit higher overall. ---Prior Discussion--- The 00z ECMWF/ECENS mean are the clear outliers in initial assessment of in the ensemble spread even compared to prior cycles of the ensemble. However, the 00z GFS has shown a trend to be slightly faster having a much less amplified (ie closed low) across the South. The 00z NAM remains the strongest and therefore, slowest to break down and stretch NE-SW in the Western Atlantic into the Eastern Gulf. Much of this variation appears to be related to the magnitude of upscale latent heat feed-back (not grid-scale feedback) from the MCS complex expected to emerge from N Mexico tonight. The greater convection and closer to the centroid of the shortwave, leads to more concentrated closed low. The 00z FV3-GFS depicts this process well compared to the operational GFS, showing strong convection near the upper low feeding a more closed upper wave and slower to weaken surface wave. Given the complex has not developed, this leads to a sizable uncertainty to this feedback. The 00z ECWMF while typically reliable, shows an atypical result and so it less favorable, while the GFS trending this way, falls in line with a typical bias but is near/closer to the UKMET. As such will favor a GFS/GFS-FV3/UKMET/CMC blend compromise at slightly below average confidence. ...Upper-level trof entering Northwest into Intermountain West Tues/Wed... ...Central High Plains Surface low Wed... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: General model blend 00z GFS/ECMWF/UKMET after 10/00z Confidence: Slightly above to above average 07z update: The ECMWF trended a bit slower and while it is still south with the QPF axis it remains close enough to continue same weighting. The CMC remained weaker but also further north, so like the NAM will remove it after 10/00z, but still remains viable across the Intermountain west. The UKMET remains in fairly good agreement with the GFS/ECMWF as a good middle ground between the two by the end of Day 3. As such a general model blend remains favored through 10/00z with removal of the NAM/CMC afterward...at high confidence. ---Prior Discussion--- Strong baroclinic zone and zonal flow supports a strong atmospheric river into the OR/CA vicinity with very good model agreement but towards late Monday, greater amplification to the flow starts upstream of the Coast building toward landfall by early Tuesday. Model agreement continues to be fairly impressive though the Intermountain west, though the ECMWF/ECENS mean start to shift a bit faster, likely due to the downstream faster evacuation across the south. This leads QPF axis/intersection with return moisture from the Gulf to develop a bit further south as it emerges into the Plains overnight Tues into Wed, compared to most of the other guidance. As such will reduce some weighting of the ECMWF late in the period. Likewise, by the end of Day 3, the 12z CMC is a bit more north and weaker with the developing closed low. To reduce its importance, will weight a bit lower too. The 00z NAM does not show much weakening across the Northern Plains of the trof showing no pinching off into a closed low, this allows for the NAM's QPF axis to shift north into MN, while other guidance would suggest otherwise. As such will eliminate the NAM in the preference after 10/00z Wed. So all in all a general model blend can be employed with only elimination of the NAM, very late and lowering of weighting to the CMC then the ECMWF slightly. Confidence is slightly above to above average in this blend given the small model variation. Model trends at www.wpc.noaa.gov/html/model2.html 500 mb forecasts at www.wpc.noaa.gov/h5pref/h5pref.shtml Gallina