Model Diagnostic Discussion NWS Weather Prediction Center College Park MD 102 PM EDT Sun Mar 18 2018 Valid Mar 18/1200 UTC thru Mar 22/0000 UTC ...See NOUS42 KWNO (ADMNFD) for the status of the upper air ingest... 12Z Model Evaluation...Including Model Preferences and Confidence ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ...Potent Shortwave Trough Rapidly Traversing CONUS from the Four Corners Today...to the Mid Atlantic Coast Tuesday... ...Surface Low Developing in the Southern Plains...Moving into Tennessee Valley Mon Ngt...Transfer to Coastal Low Tuesday... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: Blend of 12Z GFS, 00Z UKMET/ECMWF Greatest weight on the ECMWF Confidence: Average Models still show important differences with this system beyond 12Z Monday, with some run-to-run inconsistencies in addition to the model-to-model differences. The 12Z NAM continues to have the strongest depiction of the trough, holding on to a closed low at 500mb well after the other models open it into a wave as the trough axis gets east of the Mississippi River. This leads to an overall slower progression on the NAM with the associated vorticity maximum aloft situated to the northwest of other models. Its surface low position is not far off the other models, but it takes longer to complete the transfer to a coastal low. The lower heights aloft lingering over the Mid Atlantic also lead to a different trough structure as the secondary wave (discussed below) arrives. For these reasons, the NAM is not preferred at this time. Otherwise, the operational global models show reasonable similarity. In fact, most of the isoheights aloft are contained within the envelope of all the ensemble members from the GEFS, ECMWF, and CMC, which suggests that any of them could be offering a plausible scenario. One of the primary differences continues to be with the 00Z CMC which shows slightly higher heights aloft through much of the eastern CONUS in advance of the trough, and ensemble sensitivity analysis shows that the surface low intensity and track is sensitive to the strength of this ridging. Given that the other models (including the NAM) are well clustered with less ridging than is shown by the CMC, it is also not preferred. The 00Z UKMET, 12Z GFS, and 00Z ECMWF all seem plausible, so the preference is a blend of the three, with a weight toward the ECMWF as it is closest to the better ensemble clustering and the ensemble means. ...Low Amplitude Shortwave Reaching California Today...then Amplifying over Southern States Monday-Tuesday... ...Secondary Surface Cyclogenesis Along Coastal Carolinas by Tuesday Night and Possibly up the East Coast... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: Blend of 12Z GFS, 00Z ECMWF Confidence: Slightly Below Average As discussed with the first system (initial wave arriving on the East Coast), model differences in the first 24-36 hours do play important roles in how the models handle this system and the secondary cyclogenesis along the East Coast. Given that the NAM and CMC are not preferred for the initial system, they are also not preferred here. The 00Z CMC in particular seems to carry forward its bias for higher heights in showing more amplified ridging ahead of this secondary wave. Ensemble sensitivity analysis continues to show that a sharper downstream ridge contributes to a deeper surface low closer to the East Coast. Again, the CMC is within the envelope of ensemble spread, so it is not an impossible scenario, but the model preference at this time is weighted toward other global models. Examining ensemble member distribution of surface low positions, there are a considerable number of GEFS and ECMWF ensemble members that are at least slightly to the left of the ensemble mean positions (although not many as far left as the 00Z CMC). The 12Z GFS and 00Z ECMWF continued a noticeable trend of pulling the surface low track further northwest, and this is not uncommon about 2-4 days prior to the verifying event. The GFS and ECMWF are also very close to the current ensemble mean positions. Current trends and ensemble distributions therefore suggest these models have some room to nudge closer to the coast in subsequent runs. The 00Z UKMET, meanwhile, has a flatter wave overall, with a weaker surface low that is further offshore. Given that this is in contrast to the trend and the ensemble information, the preference is to lean toward a blend of the GFS and ECMWF at this time. ...Large Closed Low over the Northwest Rapidly Breaking Down Today and Tonight into Small Distinct Waves... ...Limited Motion of the Waves in Collapsing Shear Axis Mon-Tue... ...Remnant Vort Maxes Feeding into East Coast Trough Wednesday... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: General Model Blend Less weight on 12Z GFS; More on 00Z ECMWF Confidence: Slightly Below Average The chaotic breakdown of the closed low and the associated individual waves that result does not lend itself to a specific model preference. There is some variety among the different models, and even CMC, GEFS, and ECMWF ensemble means show a fair amount of spread. Therefore, a general model blend is the preference at this time. The one systematic bias relative to other models appears to be the 12Z GFS, which advances the distinct waves faster to the southeast across the Northwest and Northern Plains. Therefore, slightly less weight is placed on the GFS in this region, with a greater weight on the 00Z ECMWF. ...Closed Pacific Low near 35N/145W Drifting East and Eventually Opening into a Wave near 135W by Wednesday Morning... ...Increasing Deep SW Flow and Atmospheric River into Srn CA... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: General Model Blend; Greater Weight on 00Z ECMWF Confidence: Slightly Above Average The CMC appears to be showing its recent bias in amplifying the downstream ridge much more than other models. Isoheights at 500mb are as much as 2-3 degrees of latitude north of the GEFS and ECMWF ensemble means along the West Coast by Tuesday morning. These substantial differences would almost certainly affect the placement of precipitation and other sensible weather impacts, and thus the CMC is excluded from the preference for this system. Otherwise, the remaining models are similar enough that a general blend is preferred. The NAM and GFS do break down the closed low a bit faster than the ECMWF, and a slower transition to an open wave seems a little more reasonable given that the low is effectively cut off from a lot of other influences through Tuesday morning. So greater weight will be placed on the 00Z ECMWF. ...Shortwave Rounding Pacific Ridge over near Alaska by Tuesday... ...Digging Sharply South and Developing a Closed Low over the Gulf of Alaska by Wednesday... ...Associated Pressure Falls and Possible Surface Low Development NW of Vancouver Island... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: General Model Blend; Greater Weight on 00Z ECMWF Confidence: Slightly Below Average The NAM and the GFS may break down the closed low further south faster because they also show a faster southward progression of the height falls associated with this developing low over the Gulf of Alaska. This would exert more influence over the system to the south earlier in the forecast period. Given the strength of the closed low developing over the Gulf of Alaska -- agreed upon by all models -- the expectation is that a slower progression would also be favored here, which also argues for a lean toward the 00Z ECMWF. Given uncertainties with the initial wave (over the NW Pacific at the moment), the preference is for a general model blend at the moment as individual models can't be entirely discounted. However, a greater weight will be placed on the ECMWF for reasons stated above. Model trends at www.wpc.ncep.noaa.gov/html/model2.shtml 500 mb forecasts at www.wpc.ncep.noaa.gov/h5pref/h5pref.shtml Lamers