Model Diagnostic Discussion NWS Weather Prediction Center College Park MD 134 PM EST Thu Feb 15 2018 Valid Feb 15/1200 UTC thru Feb 19/0000 UTC ...See NOUS42 KWNO (ADMNFD) for status of the upper air ingest... ...12Z Model Evaluation with Final Preferences/Confidence... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ...Several Low-Amplitude Shortwaves Moving Quickly East through the Great Lakes and Northeast through Friday Night... ...Deepening Surface Low in Southeast Canada and Trailing Cold Front Sweeping Through the Northeast and Ohio Valley... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: General Model Blend Confidence: Slightly Above Average ---18Z UPDATE--- No change to the preliminary model preference, as model spread has continued to narrow. ---PREV. DISCUSSION--- Models do show some minor differences with this system, particularly with a secondary shortwave that pushes through the Northeast late on Friday. All models show the lobe of vorticity associated with this wave getting elongated and sheared out with time, but the 12Z NAM and 00Z UKMET show a much flatter appearance by 17/00Z, while the 12Z GFS and 00Z CMC/ECMWF show more definition. These subtle changes amount to small differences in the mass fields, but there is no clear preference for any particular small change. Therefore, the preference with this system is a general model blend. ...Shortwave Currently over the Gulf of Alaska Digging into the Northern Rockies Friday Night... ...Kicking East to the Ohio Valley By Saturday Evening... ...Associated Surface Cyclogenesis off Northeast Coast Sunday... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: 40% ECMWF, 20% EC ENS Mean, 40% Other Models Confidence: Slightly Below Average ---18Z UPDATE--- No change to the preliminary preference with this system. The 12Z ECMWF did trend a little slower and stronger with the shortwave over the Midwest, which also led to a corresponding slight northwest shift in the surface low position along the East Coast. However, the adjustments were small enough that the ECMWF still sits in the middle of the NW-SE low position spread. The 12Z CMC did shift further offshore, closer to the GFS, while the 12Z UKMET looks similar to its previous run. As model spread remains roughly similar, with models on similar ends of the distribution, the model preference remains the same. ---PREV. DISCUSSION--- Differences with this system appear to particularly relate to the strength and timing of the shortwave as it pushes through the Midwest on Saturday. The models that have a slower and stronger shortwave (12Z NAM and 00Z UKMET) tend to have a stronger surface low off the Northeast coast that is positioned further to the Northeast. Meanwhile, the models with a faster and slightly weaker shortwave (12Z GFS, 00Z ECMWF/CMC) have a surface low that is further offshore of the Northeast coast. An examination of ensemble low positions over the past four model cycles shows that although SW-NE displacement (related to timing) has narrowed, the NW-SE spread is roughly the same as it was a couple days ago. Adding complexity is the fact that the shortwave that reaches the East Coast will likely result from some phasing of the northern stream with the lingering trough over the Southwest. This may result in continued uncertainty for several more model cycles. Models also show indications that the low-level cyclone will be influenced by diabatic heating effects, with the GFS and ECMWF showing a ramp-up in low-mid level PV as the surface low develops over the Atlantic. The GFS is perhaps more aggressive with this, particularly initially (Saturday Night) and ties the surface low very closely to the QPF max, which eventually results in a surface low on the southeastern extent of ensemble low positions. The model preference for this system, therefore, leans toward a greater weight on the 00Z ECMWF which sits closest to the ensemble means and roughly in the middle of the NW-SE low position spread off the East Coast. Regarding the shortwave upstream, all the available models sit well within the ensemble spread for upper level heights, so any of them could be reasonable possibilities. Therefore, a general blend is preferred with a 40% weight on the operational ECMWF, 20% on the ECMWF Ensemble Mean, and 40% on the remaining models. ...Next Trough Digging into the Northwest this Weekend... ...Surface Low and Front in the Intermountain West with Lee Cyclogenesis in the Plains by Sunday Evening... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: Non-UKMET Blend Confidence: Average ---18Z UPDATE--- No change to the preliminary preference. ---PREV. DISCUSSION--- Model spread with the height fields aloft has continued to gradually narrow over the past several model cycles, although some spread still exists -- especially with the amplitude of the trough. Although spread remains, there is no clear preference for one operational model over the other, and all models fit within the envelope of ensemble scenarios. The one exception is with the 00Z UKMET which has lower heights over the interior Northwest than most of the other models, and the more expansive trough may have an effect on sensible weather. Model trends at www.wpc.ncep.noaa.gov/html/model2.shtml 500 mb forecasts at www.wpc.ncep.noaa.gov/h5pref/h5pref.shtml Lamers