Model Diagnostic Discussion NWS Weather Prediction Center College Park MD 251 AM EDT Sun Mar 18 2018 Valid Mar 18/0000 UTC thru Mar 21/1200 UTC ...See NOUS42 KWNO (ADMNFD) for the status of the upper air ingest... 00Z Model Evaluation...Including Model Preferences and Confidence ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ...Persistent W-NW flow aloft over Northeast US through Monday... ...Weak surface low tracking from Lake Superior today to Long Island Sunday associated with digging shortwave aloft... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: General model blend Confidence: Above Average Models remain in excellent agreement in mass fields from Hudson Bay through the Northeast US and Southeast Canada in the early periods of the forecast with the main cold air advection surges and associated shortwaves crossing through. This includes the last stronger wave on Monday that is responsible for the weak surface wave moving from Lake Superior to the Hudson Canyon region off NY/NJ by early Monday. The model break from consensus occurs with the remaining baggy trof over the E Ontario to Nova Scotia through Tues/Wed. The 12z CMC is generally weaker and quicker to shift the weaker wave out with the main closed low, which is further SW over the Gulf of St. Lawrence towards NB) by Tuesday than the ensemble solutions support. The 12z UKMET which is very strong with the inner core of the same closed low rotates back through the Gulf of St. Lawrence Tuesday as well... though it handles the remaining sharpening of the lingering baggy trof over northern New England by Wed matching the 00z NAM, 12z ECMWF and ensemble suite, potentially making it useful over the US domain. The 00z GFS is a bit weaker with the eventual baggy trof sharpening presented by the NAM/ECMWF/UKMET on WED, as it shifts the energy downstream a bit quicker but has trended toward the NAM/ECMWF/UKMET. All considered a non-CMC blend is preferred for the US domain in the Northeast...though larger scale less of the UKMET would be preferred in SE Canada. Confidence is above average given the overall differences are small and the impacts on sensible weather are low. 07z update: The 00z CMC shifted toward common agreement with the other guidance to support a general model blend at above average confidence. ...Deamplifying wave becoming increasingly sheared Out in central Appalachians and through Carolinas later today with associated surface low... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: General model blend Confidence: Above average There is very little model spread with the mass fields for this system...and no significant differences were noted with the 00z NAM/GFS runs to break from a general model blend preference at above average confidence. ...Potent Shortwave at the Base of Devolving Western Closed Low...closes off while ejecting into Plains late Sunday with strong lee cyclogenesis. Low will press through to TN Valley by early Tues and transfer to coastal low late Tues... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: 00z ECMWF/UKMET/GEFS blend Confidence: Average GOES-WV mosaic depict in S CA a very strong shortwave at the base of closed low that dominates the West, as well a low amplitude shortwave emerging out of the Subtropical Eastern Pacific at Baja CA at this time. The Subtropical wave will lift north into TX and aid negative tilt to the main northern stream shortwave out of NM and rapidly develop a strong lee cyclone late Sunday into Monday and start to shift with good pace across the southern Plains toward the MS River Valley by Monday. It is at this point, deterministic guidance begins to differ in handling the system as it moves into generally confluent upper level flow and weakens. The first to depart is the 12z CMC which is a bit slower and stronger through the TN valley and very slow/weak in the transfer of energy to the developing coastal low midday Tues. The UKMET shows some typical bias of being too strong/slow to spin down the wave into the confluent flow though this counteracts its tendency to be fast and so paces extremely well with the remaining guidance and ensemble suite but is a clear stronger wave (nearly 4mb for the surface low by 18z Tues. The 00z GFS is well timed with the 00z NAM with respect to the lead 5H wave/sfc low but the 7H mass fields belay some strength/upstream interaction influences that suggest the 00z NAM may be a bit too aggressive. While there are some smaller scale detail differences by day 3 with the 12z ECMWF and the GFS but this seems minor and within reason/ensemble spread. Additionally, the strengthening upstream shortwave across the Gulf coast as well as sharpening of the larger scale trof (as energy from the northern plains slides southeast) by day 3 lead to further reduction in predictability/certainty... please see sections below for these specific waves. As such 00z GFS and 12z ECMWF blend is preferred at average confidence. 07z update: While the 00z CMC trended favorably toward a weaker solution moving through the TN valley, there was broad cyclonic rotation up the Appalachians instead of across and through the southern Mid-Atlantic states Tuesday. The 00z UKMET and ECMWF both slowed slightly and while following a similar track/evolution to the 00z GFS there has been a increase in timing difference between the 00z GFS and the ECMWF/UKMET. This is in line with the high sensitivity and while cannot fully discount either fast or slow timing, this trend toward slower and stronger upstream wave has been growing with this shift. In fact, the 00z GEFS mean has also slowed and the mean surface wave is better clustered to these solution than the operational 00z run. As such will favor the 00z ECMWF/UKMET with the GEFS mean replacing the operational GFS in the preference. ...Low amplitude shortwave reaches California Sunday amplifying into secondary trough Mon-Tue over the Mid-South reaching the Southeast late Tuesday supporting secondary coastal low early Wed off SC Coast ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: 00z GFS/ECMWF blend Confidence: Slightly below average In the wake of the main shortwave in the southwest today, a piece of shortwave energy is shed from the closed low in the central Pacific through the ridge that separates the larger features and enters CA on early Monday with little fanfare and continues to be low amplitude through early Tuesday into the Red River valley. Here the larger scale weak trof exists and with stronger northwesterly jet flow aloft, it can amplify across the MS Valley and slow through the Eastern Gulf states. Along with northern stream shortwave energy descending from the northern Plains (see section below) on Tuesday/Wed, generate a stronger neutral to weak negative tilt and support a secondary surface low off the SC coast that lifts northeast by the end of the forecast. The 12z CMC having lagged with a stronger more closed mid-level circulation through the TN valley is out of place from the onset and shows a deeper latitude response to troughing as a whole. The 00z NAM similarly has a deeper latitudinal trof having greater amplification of the wave through the MS Valley and northern stream support to lead to a closed 5H feature in the TN valley by Wed...a bit too fast/too strong. The 00z GFS gets to a similar point but is much softer and weaker with the wave and its binary interactions with the other upstream features. While the 12z ECMWF is slower than the GFS, its evolution is quite similar. The 12z UKMET is a bit faster/stronger with this wave typical of its bias but not dramatically so to completely discount it; it is discounted mainly due to the lack of upstream digging trof. All other guidance/ensemble support the northern Plains system to descend, but it does not in the 12z UKMET. As such a 00z GFS/12z ECMWF blend is preferred. Confidence is good, but given the moving parts/influences of downstream setup as well as upstream timing leads to slightly below average confidence due to high sensitivity. 07z update: The 00z ECMWF trended a bit more amplified/faster at the top of the trof in the southern Appalachians on Wed enhancing downstream shortwave ridging and pivoting the surface low a bit closer to the coast by midday Wed matching very nicely with the 12z ECENS mean and 00z GEFS mean...though not as far left as the NAM or 00z CMC. The UKMET is well out to sea respective to the cluster as it appears to be melding with the upstream wave (section below) and not allowing a slowing/pivot to develop. As such a 00z ECMWF/GFS remains preferred at slightly below average confidence. ...Large Closed Low rapidly breaks down Sunday into smaller-scale distinct waves with several becoming increasingly embedded in NW Flow over the Northwest and Northern Plains as northern stream jet sags south, eventually expanding in broad negative tilt trof over southern third of US Wed... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: 00z GFS/ECMWF blend Confidence: Slightly below Average GOES-WV mosaic depicts shortwave emerging from the Beaverhead range into SW MT and a elongated wave over the SW Prairies of Canada, these two waves will interact across S Canada and as the mean Arctic jet streak shifts east over Manitoba; the northern stream is able to buckle and meld with the dancing shortwave pair leading to weak broadening troughing across S Saskatchewan by late Monday into Tuesday. Additional lingering energy through the Rockies and the subtle Pacific shortwave (see section above) will allow for this wave to descend into the central Northern Plains and stretch southeast eventually carving out a larger scale trough over the eastern US by Wed. The 12z UKMET is the main outlier with this interaction having been a bit stronger and shifting east more. This enables the tail end of the trof over Hudson bay to influence it dragging the weak wave eastward and not descend into the central Plains...breaking significantly from run to run ensemble trend/continuity and other deterministic guidance. The 00z NAM shows stronger amplification/winding up the inner core of the trof over Canada slipping into N MN by 12z Wed, falling into a typical Day 3 negative bias for the model. In this flow regime strong northwesterly jet from the Pacific should keep the shortwave energy stretched out in carving out the western portion of the trof in the MS Valley, but the 12z CMC does not do this preferring to focus on more consolidated waves in the northern Plains/MN by Wed...so here too the model falls into negative bias. This leaves the 00z GFS which also shows typical fast bias descending into the western portion of the trof by Wed, but the 12z ECMWF is generally equally slow...so here a nice typical compromise between the 00z GFS/12z ECMWF is likely a solid bet. Confidence is slightly below average given high sensitivity of the environment and likely run to run variability even though the spread between the GFS/ECMWF is small and agreeable in the ensemble suite. 07z update: The 00z UKMET trended more favorably with weaker distinct waves initially and less interaction with return shortwave from NE Canada...still the Pacific jet is a bit stronger and faster enhancing the digging of the shortwave on the western side of the developing larger scale trof by Wed. This leads to a flatter broader base to the trof across the Southern US by the end of the forecast. The 00z ECMWF trended favorably toward the GFS with a slightly faster southeastward digging on day 3. With downstream issues the 00z CMC is not favorable for this section of the trof as well. As such a 00z GFS/ECMWF blend remains favored at slightly below average confidence. ...Closed Low settling south in the Pacific to around 35N/145W by Sunday afternoon and slowly drifting east through Tuesday increasing southwesterly flow with possible atmospheric river approaching California by Tuesday night... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Preference: 00z UKMET/ECMWF/CMC blend Confidence: Average Stationary closed low remains between 130-140W and remains until about Tuesday when an imbalance within the inner core of the low manifests. This in combination with weak remaining smaller-scale remnant shortwave energy from the weakening Western US upper low along the OR/WA coast, appear to draw out some model differences. The 00z GFS has backed off the stronger internal wave in the closed low but remains most amplified/influential with the s/w along the Pacific coast leading to the closed low to shift east by the end of the forecast period. While this is supported by the NAM and GEFS members, this looks less likely compared to more concentric and weaker waves presented by the 12z ECMWF/UKMET. The 12z CMC may be too far elongated...shifted SW with the closed low on Wed to support it in the blend. Either way...strong southwesterly flow ahead of the closed low and associated surface wave should support strong atmospheric river directed toward central or southern CA toward the end of day 3 into 4. Given the issues with the GFS...WPC prefers the slower ECMWF/UKMET timing with respect to the rainfall as well. Confidence is average. 07z update: The 00z UKMET/ECMWF and CMC all resolve a stronger interaction with the southward sinking wave allowing for additional eastward propagation relative to the 12z run. While this is an adjustment toward the NAM/GFS, the 00z GEFS mean is still south and west of both the GFS/NAM and nearing the ECMWF/CMC and quite similar to the UKMET. Overall a blend of the UKMET and the ECMWF/CMC seems a good compromise to adjust for these shifts. Confidence remains average. Model trends at www.wpc.ncep.noaa.gov/html/model2.shtml 500 mb forecasts at www.wpc.ncep.noaa.gov/h5pref/h5pref.shtml Gallina