Model Diagnostic Discussion NWS Weather Prediction Center College Park MD 1220 PM EDT Tue Sep 03 2019 Valid Sep 03/1200 UTC thru Sep 07/0000 UTC ...See NOUS42 KWNO (ADMNFD) for the status of the upper air ingest... 12Z Model Evaluation, Confidence, and Preferences ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ...Overall Pattern Across the CONUS... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: General model blend Exception: 12z NAM/GFS and 00z ECMWF over North Pacific after 06/0z Confidence: Slightly above average Average across Great Lakes end of D3 Broad global trof anchored in S Hudson Bay continues slow shift north and east, especially after embedded shortwave and associated deep surface wave cross through the Great Lakes today into SE Canada by early Thurs with cold front pressing through New England. This is well agreed upon, with exception of the 12z NAM still lingering a deeper trof through the Mid-Atlantic around 60hrs...this is likely leading to the NAM breaking away from the consensus with Dorian after this time period. Afterward, the strong Gulf of AK jet and closed low/shortwave entering SW Canada currently, will interact but lead to fairly flat Northwesterly flow across Central Canada. The Pacific shortwave will lead the nose of the 250mb jet, too far and shear/weaken into the NW Great Lakes by late Thursday. Typical alignment occurs with this wave with the ECMWF/CMC slower/lagging. However, by shortening the wave length to the nose of the jet, this leads the ECMWF/CMC to be faster with the next wave and slightly dampened into the confluent flow relative to the 12z NAM/GFS and UKMET. These are very subtle timing differences and are very difficult to lock down, as noted by the run to run variation. Admittedly, this has high impact to the track of Dorian as well. Both the NAM/GFS are slightly stronger with the upper jet making the ECMWF look of of sorts, still this is lower impact to QPF/sensible weather across the Northern Tier into the Great lakes to support a general blend here (see proxy to NHC for Dorian below). Across the Northwest, however, there is sufficient spread but not ashore even through 84hrs. The remaining closed low that is offshore by Thursday, continues to lift/shear into the Rockies ridge breaking it down a bit by Friday. The 12z NAM/ECMWF are slightly stronger by 84hrs in line with their biases (EC-slow and NAM-strong by end of D3) but with little negative affect for QPF across the Rockies. The spread, is the deep latitude trof approaching out of the Gulf of AK, that kicks the precursory wave into the ridge, and the 00z UKMET is a clear fast outlier, early. Eventually, the CMC is stronger, more compact further north and shifts eastward with the trof axis. This also makes it less preferable, so would shift to a 12z NAM/GFS and 00z ECMWF blend in the Pacific for the end of D2 into D3. Though elsewhere, a general model blend can be employed with slightly above average confidence (average over the Great Lakes into D3 given run to run variance and critical importance of timing) ...Hurricane Dorian... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Please refer to the official National Hurricane Center forecast track Initially, as Dorian tracks NNW, the guidance is fairly well clustered to the official NHC track. But as it turns and lifts parallel to the Carolina coasts, timing differences begin to manifest with the 12z NAM continuing to accelerate well out of place of the otherwise solid consensus. However, the 00z ECMWF eventually breaks too slow after 60hrs as well...that both are not in a best proxy. The UKMET redevelops much deeper and slows relative to the official track forecast about 72-84hrs; the 12z GFS slowed a bit as well driven by a stronger jet axis over the Appalachians for a longer period of time even with the approach of the northern stream shortwave/jet energy. This seems quite favorable but slows the track of Dorian off the official track from NHC. So the 06z GFS and 00z CMC are closest in proxy to the track... but inclusion of the 00z UKMET and 12z GFS seems a good idea given the approach of the shortwave/baroclinic influences toward 84hrs. ...Tropical Depression Seven.... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Please refer to the official National Hurricane Center forecast track NHC started classifying the West Gulf system at 09z. The 15z of TD Seven is forecast along the northern periphery of the main deterministic global guidance, closest to the 12z NAM and 00z UKMET. However, the NAM has a northward bias with the extending the 850-7H trof bringing QPF much further north than expected, opposed by the 12z NAM-CONEST, which is much closer to the WPC preference with respect to QPF in TX. The 12z GFS shows similar northward shift compared to 00z run and just a few miles south of the official track (especially through landfall). So for winds/mass...the best proxy is the 12z NAM-CONEST/GFS and 00z UKMET for the 15z forecast, however, the NAM may work in the mass fields, if one does not have the CONEST to blend. Model trends at www.wpc.noaa.gov/html/model2.html 500 mb forecasts at www.wpc.noaa.gov/h5pref/h5pref.shtml Gallina