Model Diagnostic Discussion NWS Weather Prediction Center College Park MD 141 AM EST Sat Jan 26 2019 Valid Jan 26/0000 UTC thru Jan 29/1200 UTC ...See NOUS42 KWNO (ADMNFD) for the status of the upper air ingest... 12Z Model Evaluation...with Preferences and Confidence ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ...Gulf of Mexico/Florida... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: 00z GFS/UKMET/CMC blend Confidence: Average 07z update: The 00z ECMWF trended a bit faster and less amplified but remained slower nearly matching the 00z NAM. Given the slowing/amplification of the CMC and a bit of increased amplification of the UKMET all shifting to the general timing of the initial preference (ECENS/GEFS mean) there is increased confidence to hedge this way even further supporting a 00z GFS/UKMET/CMC blend eliminating the use of the means to better pull out the signal of the wave as a whole. Confidence is now average in this preference. ---Prior Discussion--- Highly amplified pattern with deep ridge along the West Coast up into S AK and large global-scale vortex centered near Hudson Bay keeps synoptic agreement very high through the bulk of the short-range forecast (though by Tuesday the timing of deep cold blast is moderate across Manitoba to N MN). Model spread continues mainly with the smaller scale shortwave features moving around the edges of the vortex as a whole, with differences mainly in timing of the wave reaching key amplification points, such as the base of the trof, phasing with other waves or over-topping the ridge in AK/Yukon. The first, and highest model variance continues to be at the furthest edge of the of trof as the shortwave currently in S AZ drops through to far NE Mexico and its influence/amplification of a surface wave along the tropical moisture stream in the W Caribbean through the Straits of Florida. The 12z ECMWF continues to be slowest with greatest phasing/merging of the shortwave across the Northern Gulf, slowing even further before shifting across FL supporting a deeper surface wave. The 18z and 00z NAM shifted very heavily toward this solution but both remain on the western/slowest example of the ECENS members which while more amplified than CMCE/GEFS members, are flatter than the ECMWF/NAM. The 00z GFS like prior members, continue to slow and show more amplification but still remain more flat and even faster than the 18z GEFS mean (which is closer to the ECENS mean. The 12z CMC is now the fastest and flattest with the wave, while the UKMET is slower but also quite flat. Thinking this slow shift toward increased amplification is correct (likely due to some latent heat aiding wave amplification but given no significant downstream blocking, ECMWF/NAM solutions do not seem plausible give fast flow, advection of said latent heat feed-back. As such favor the 00z GFS/12z UKMET and 18z GEFS/12z ECENS mean blend. Confidence is slightly below average for this wave. ...Northern Plains/Midwest/Great Lakes... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: General model blend Confidence: Above average 07z update: A much tighter clustering as set up as the 00z CMC/UKMET both track very closely to the 00z GFS/NAM. Even the 00z ECMWF made a sizable shift eastward with the wave into far Northern Canada, while still slower/west than the GFS/NAM it is much closer and now the deformation zone/surface cyclone track shifted much more north (in line with 12z CMC/UKMET). Will continue to support a general model blend but would now include the arithmetic blend to the spatial one suggested before...as such confidence is also increased to above average for this system...though the cold air/main upper level closed low remains different in timing and hedging more GFS for that timing. ---Prior Discussion--- Further north, first pocket of dry cold air, produce a highly agreeable surface wave/deep cyclone through the N Great Lakes by Sunday; while there is some timing differences (GFS fast, CMC a bit slow, both quite occluded in E Quebec by Mon...there are minor affects to the approaching deeper shortwave by the beginning of the week. This next wave, is still well out to sea out of the deepest observation network but still is timed well over-topping the ridge into S AK, Yukon by later today into Sunday. Once in northwest flow, slow development occurs and really starts to deepen across the Upper Midwest. The model differences appear to be completely driven by timing of the energy over the ridge. The faster solutions of the 00z GFS/FV3 and NAM hence place the energy further east and given jet/cold air placement, frontogenetic forcing aids a deeper faster surface wave, that reaches the base of the trof across S WI into LP of MI by Mon into Tuesday. The slower ECMWF, takes the bulk/center of the energy/jet focus more south along the spine of the Rockies and leads to a slower, southern track with the surface wave. The evolution/shape/timing (relative to the NW flow) are pretty similar so this adds some overall confidence. The UKMET/CMC are more central between the other solutions with the UKMET faster compared to the CMC. Here a general model blend can be supported but not a typical arithmetic blend... but one that recenter the features to a more common center...ie keeping the overall magnitude of the surface low, QPF, winds and other sensible wx fields into a more common center point. Confidence is slightly above average in this common center blend. Additionally, the secondary cold push/arctic air mass should be more central as well...closer to the GFS/NAM center of cold air in S Manitoba by 12z Tuesday and not as fast as the UKMET or slow as the ECMWF. Model trends at www.wpc.ncep.noaa.gov/html/model2.shtml 500 mb forecasts at www.wpc.ncep.noaa.gov/h5pref/h5pref.shtml Gallina