Hi Everyone, Friday map discussion for 22 Nov 2013 opened with a brief discussion of several recent examples of significant ensemble and deterministic forecast errors on 5-8 day times scales, moved on to an overview of the Midwest severe weather outbreak of 17 Nov 2013 (http://www.cnn.com/video/?/video/bestoftv/2013/11/21/cnn-drone-surveys-illinois-tornado-damage.cnn&video_referrer=; http://www.atmos.albany.edu/student/ppapin/lb13_img/special/20131118/images/radar.html), was followed by a more extensive discussion of the Mediterranean cyclogenesis and heavy rain event (Sardinia) of 17-19 Nov (http://www.washingtonpost.com/world/europe/apocalyptic-storm-floods-parts-of-sardinia-downing-bridges-and-killing-at-least-14/2013/11/19/91be1bfa-50eb-11e3-9ee6-2580086d8254_story.html), and concluded with a brief discussion of the phasing-related forecast uncertainties associated with a possible storm along the East Coast on Wed 27 Nov that had the potential to disrupt pre-Thanksgiving holiday travel. The master discussion web link can be found here: http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/. Alicia Bentley, Chip Helms, Julian Quinting, Ben Moore and Philippe Papin prepared an assortment of loops and imagery for the discussion and helped to identify key science issues that were addressed during the discussion. The examples of forecast error that we discussed briefly included the ensemble forecast bust of a North Pacific blocking regime on 1200 UTC 12 Nov (subject of an earlier map exchange initiated by Tony Fracasso; ), the false alarm deterministic ECMWF "fantasy East Coast cyclone" forecast bust in the 192 h forecast verifying 0000 UTC 15 Nov (http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/euro.html), forecast uncertainties associated with the configuration of the large-scale flow pattern over western and central North America on 1200 UTC 17 Nov immediately preceding the late season severe weather outbreak over the Midwest later that day, and the strength and location of a 500 hPa cutoff cyclone located over the Southwest at 0000 UTC 22 Nov that was destined to play a role in a potential significant east coast cyclone on 26-27 Nov before the Thanksgiving holiday. 1. North Pacific Blocking Forecast Uncertainty of 1200 UTC 12 Nov 2013: Fracasso posted images that showed "a rather 'impressive' failure [verifying 1200 UTC 12 Nov] on the part of both the GFS and ECMWF ensembles that was illustrated by his 'outside the envelope graphic for the GFS/ECMWF/CMC ensemble members, their means, and the deterministic ruins." (http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/map.txt; http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/map/ensemble_1.gif; http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/ensemble_bust.html) Fracasso indicated that "there were many members from both ensemble systems that indicated energy would tuck back southwestward under the ridge in the Pacific... thereby holding back a wholesale trough that was forecast to move into the West Coast... there are zero members that show strong ridging along 110W and the troughing moving through the Great Lakes." The major forecast error was that the various models forecast a Rex block to develop over the central North Pacific with a deep trough forecast to form over western North America. In reality, an omega block formed instead with a trough downstream over the eastern North Pacific and a large amplitude ridge over western North America (http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/map/ensemble_4.gif; http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/map/ensemble_5.gif). The transition from the incorrect forecast scenario over the North Pacific, Rex block, to the correct observed scenario, Omega block, was fairly abrupt in the GFS (between 168-180 h for forecasts verifying 1200 UTC 12 Nov)(http://www.atmos.albany.edu/student/abentley/mapdisco/20131122/pacific.html). The hints are there as to where look for sources of model forecast uncertainty (e.g., what physical processes governed the nature of ridge-trough interactions in an environment of broad anticyclonic flow and relatively low absolute vorticity at high latitudes at time scales beyond 168 h and how did these processes change at shorter time scales?). Preliminary indications are that forecast uncertainties associated with poleward-moving cyclones along the western flank of the ridge over the central North Pacific likely impacted the strength and amplitude of the central North Pacific ridge and contributed to uncertainties in to what extent troughs dropping southward to the east of the ridge would "cut back" underneath the ridge and form a Rex block. From a scientific perspective, revisiting the work that Mullen (1987, 1988) did on Pacific blocking and explosive oceanic cyclogenesis but using higher-resolution modern gridded datasets to compute eddy fluxes of vorticity, PV, momentum, heat and moisture would likely be illuminating (http://www.atmos.albany.edu/student/abentley/mapdisco/20131122/pacific.html). 2. ECMWF Deterministic Forecast Bust of an East Coast Cyclone on 0000 UTC 15 Nov 2013: The 192 h ECMWF deterministic model forecast bust verifying 0000 UTC 15 Nov featured a moderately intense coastal cyclone situated near the coast of New Jersey that was not supported by the EPS. The deterministic ECMWF forecast solution was outside and to the left of the solution phase space defined by the EPS members. A potentially distinguishing feature of this 192 h deterministic ECMWF forecast bust was an erroneously strong forecast cyclonic circulation in the lower troposphere and a forecast vertical structure that favor deepening of the forecast cyclone in contrast to the larger scale confluent flow pattern derived from the the EPS individual ensemble members that favored a more eastward-moving and weaker cyclone (http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/euro.html). 3. Midwest Severe Weather Outbreak of 17 Nov 2013: Kyle Griffin's d(prog)/dt loops for 1200 UTC 17 Nov, the morning of the very recent Midwest severe weather outbreak, show that the deterministic GFS run did not forecast a large-scale flow configuration favorable for a "second season" severe weather outbreak until the 108 h forecast verifying this time and didn't get important details correct until the 90 h forecast verifying this time. At the longer time ranges (beyond 108 h), the Deterministic GFS runs "cut back" too much of a 500 hPa trough and the associated cyclonic vorticity into the Southwest (http://www.atmos.albany.edu/student/abentley/mapdisco/20131122/northamer.html). This error pattern resulted in too much of a confluent flow pattern downstream that was considerably less favorable for a severe weather outbreak. The critical deterministic GFS forecast error beyond the 108 h forecasts verifying 1200 UTC 17 Nov was that the model tried to "cut back" troughs over the southwest beneath the ridge over the eastern Pacific in a smaller scale version of the same error pattern seen in longer time scale forecasts over the central North Pacific in forecasts verifying 1200 UTC 12 Nov as discussed above (http://www.atmos.albany.edu/student/ppapin/lb13_img/special/20131118/images/radar.html). As for the severe weather outbreak itself, Kyle Griffin's loop of standardized 250 hPa geopotential height anomalies shows negative anomalies between -1.5 and -2.0 over the northern Plains and positive anomalies of a similar magnitude over the Southeast and Gulf of Mexico that are consistent with an anomalously strong subtropical jet over the central U.S. Corresponding standardized 850 hPa geopotential height anomalies of -2.0 to 12.5 over the Northern Plains and +1.5 to +2.0 over Atlantic Canada favored a poleward-directed low-level flow over the Mississippi Valley. Precipitable water (PW) standardized anomalies within this poleward-directed 850 hPa flow were in the +2 to +3 range. Anomalously strong shear and PW values were sufficient to offset comparatively modest CAPE values (1000 j/kg or less) in the severe weather corridor in support of the observed severe weather outbreak which was well forecast by the SPC folks beginning approximately 3-4 days in advance of the event. 4. Southwest U.S. Cutoff Cyclone of 22 Nov 2013: As for the antecedent conditions to next week's possible pre-Tahnksgiving eastern U.S. storm, Kyle Griffin's d(prog)dt loops show nicely that the 500 hPa geopotential height cutoff cyclone situated near San Diego at 1200 UTC 22 Nov was not well forecast until between 120-132 h by the deterministic GFS forecast. GFS forecasts at longer ranges diode not have a 500 hPa cutoff cyclone and did not "cut back" a trough and its associated vorticity underneath a ridge just offshore of the Pacific Northwest. So, without trying very hard we discussed multiple examples of model forecast failures on the 5-8 day time scale associated with either too much or too little "cutting back" of 500 hPa troughs and their associated vorticity beneath upper-level ridges on different spatial scales. How do you spell a forecast uncertainty problem in need of further scientific analysis (http://www.atmos.albany.edu/student/abentley/mapdisco/20131122/northamer.html)? The forecast uncertainties associated with the cutoff cyclone over the Southwest were also discussed in the conjunction with forecast uncertainties for the possible (as of Fri afternoon) eastern U.S. cyclogenesis event on the Tu-Wed before Thanksgiving. As of the 1200 UTC forecast cycle from 22 Nov both the GFS and ECMWF deterministic and ensemble forecast runs were showing considerable run-to-run temporal and spatial uncertainty in whether the ejecting cutoff cyclone from the Southwest in the southern branch of the westerlies would phase with disturbances embedded in the northern branch of the westerlies. 5. Mediterranean Cyclone and Heavy Rains of 17-19 Nov 2013: Most of the Friday map discussion was devoted to an analysis of the Mediterranean cyclogenesis event of 17-19 Nov 2013 and the exceptionally heavy rains that fell over easterly flow favored regions of Sardinia (http://www.washingtonpost.com/world/europe/apocalyptic-storm-floods-parts-of-sardinia-downing-bridges-and-killing-at-least-14/2013/11/19/91be1bfa-50eb-11e3-9ee6-2580086d8254_story.html). Anticyclonic wave breaking (AWB) and cyclonic wave breaking (CWB) played a role in this event. As is demonstrated in Alicia's loops (http://www.atmos.albany.edu/student/abentley/mapdisco/20131122/atlantic.html), CWB over the north-central Atlantic on 6-7 Nov was followed by failed AWB over the western Atlantic on 7-8 Nov. This failed AWB was followed in turn by successful AWB over the eastern Atlantic and western Europe on 9-11 Nov, downstream of which a (first) deep trough was manifest as a PV tail that thinned and fractured, and produced a cutoff cyclone over the western Mediterranean Sea on 11-12 Nov. This cutoff cyclone helped to initiate a stormy weather pattern over the western and central Mediterranean Sea. Meanwhile, a "monster AWB" event that occurred over the eastern Atlantic and northwestern Europe on 15-16 Nov (note the frontal cyclone that formed in the equatorward entrance region of the 180 kt jet on the DT at 0000 UTC 16 Nov and then reached the coast of Norway by 0000 UTC 17 Nov) set the stage for additional Mediterranean Sea cyclogenesis Downstream of this "monster" AWB event, a second fractured PV anomaly set up shop over the Iberian Peninsula by 1200 UTC 16 Nov. This second fractured PV anomaly was associated with the heavy rains over Sardinia on 17-19 Nov. This second PV anomaly first moved westward during the PV fracture process, subsequently stalled near Gibraltar, and then moved eastward across the western Mediterranean Sea. The onset of the eastward movement of the second fractured PV anomaly after 1200 UTC 17 Nov coincided with the approach of a strong polar trough from the northwest (downstream of a modest AWB event over southern Greenland) and a second, weaker trough embedded in the subtropical jet to the west of northwest Africa. The eastward movement of this second trough to the west appeared to be aided by lower-latitude AWB downstream of subtropical cyclone Melissa over the west-central Atlantic. On 18-19 Nov this second fractured PV anomaly accelerated eastward across the Mediterranean as it began to be influenced by the aforementioned strong polar trough dropping south-southeastward over the eastern Atlantic and western Europe downstream of an uber AWB event over the extreme northwestern Atlantic on 19 Nov. Additionally, an AWB event that occurred over central Europe on 19-20 Nov seems to have been induced by downstream baroclinic development in response to the aforementioned upstream AWB event. The smaller-scale details of the eastward passage of the fractured PV anomaly across the western and central Mediterranean Sea are also of interest. Visible/IR/WV satellite imagery from the CIMSS satellite blog (http://cimss.ssec.wisc.edu/goes/blog/wp-content/uploads/2013/11/METEOSAT10_15-20NOV2013loop.gif; http://cimss.ssec.wisc.edu/goes/blog/wp-content/uploads/2013/11/METEOSAT10_WV_15-20NOV2013loop.gif; http://cimss.ssec.wisc.edu/goes/blog/wp-content/uploads/2013/11/METEOSAT10_HR18NOV2013loop.gif; http://cimss.ssec.wisc.edu/goes/blog/wp-content/uploads/2013/11/METEOSAT10_HRVIS_18NOV2013loop.gif) shows several smaller-scale PV disturbances rotating cyclonically around the upper-level PV anomaly. One of these PV disturbances appears to rotate across extreme northwestern Africa and the western and central Mediterranean Sea between 0000-2100 UTC 18 Nov during which time exceptionally heavy rainfall occurred over the eastward-facing hilly terrain over eastern and northeastern Sardinia as seen in this combined rainfall and IR satellite loop (http://sop.hymex.org/radar.php?current=20131118&nav=ItalyCumulRadars&expected=SRIcomposite_Italy_NAN_A92). Loops produced by Ben Moore and Julian Quinting of 850 hPa geopotential heights, PW, and 1000-700 hPa integrated WV transport (IVT)(http://www.atmos.albany.edu/student/bmoore/IVT_Sardinia_Nov2013.html), CAPE, SLP, and 10-m winds (http://www.atmos.albany.edu/student/bmoore/CAPE_Sardinia_Nov2013.html), 48 h backward trajectories for selected variables (use the up and down arrows to see the different variables) ending at 1200 UTC 17 to 0000 UTC 19 Nov at 6 h intervals (http://www.atmos.albany.edu/student/bmoore/back_traj_sardinia.html), and cross sections along 40 N of equivalent potential temperature, zonal wind component, and vertical motion (http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/cross.html) illustrate important mesoscale details of the heavy flood-producing rainstorm that impacted portions of Sardinia (GFS 0.5 deg gridded datasets were used to produce these loops). On 16 Nov, a corridor of northeastwards-directed IVT develops over northwestern Africa around a developing 850 hPa cyclone located over the extreme western Mediterranean Sea. By 1800 UTC 17 Nov, a second 850 hPa cyclone forms over the western Mediterranean Sea while a trough develops to the south over northwestern Africa (see aforementioned satellite loop discussion). In response, two confluent IVT streams develop and merge immediately to the east of the second 850 hPa cyclone in a region of increasing PW to the west of Sardinia. This cyclone, its associated trough, and IVT transport strengthen through 1800 UTC as they approach and cross Sardinia. The corresponding loop of SLP, 10-m winds and CAPE (http://www.atmos.albany.edu/student/bmoore/CAPE_Sardinia_Nov2013.html) shows the strengthening southeasterly (upslope) flow (~15 m/s) directed toward Sardinia. This loop also suggests that a narrow band of CAPE values between 1000-1500 J/kg oriented northwest-southeast and situated just ahead of the surface trough crosses Sardinia between 1200-1800 UTC 18 Nov. This band of higher CAPE values organizes and moves eastward between 1800 UTC 17 Nov and 0600 UTC 18 Nov, likely in response to cooling aloft ahead of the eastward-moving upper-level PV anomaly. The 48 h backward trajectory analysis from 1200 and 1800 UTC 18 Nov shows the confluence of multiple airstreams above Sardinia (the anticyclonic turning of the trajectories with decreasing pressure is indicative of deep warm-air advection) (http://www.atmos.albany.edu/student/bmoore/back_traj_sardinia.html). Ascent is indicated over the last 6-18 h along trajectories that end between 0600-1800 UTC 18 Nov. Equivalent potential temperatures in the low-level air stream arriving over Sardinia from the southeast are in the 312-318 K range, hardly indicative of a tropical system. Finally, vertical cross sections along 40 N of vertical motion (shaded, cool colors connote ascent), equivalent potential temperature (black contours), and the zonal component of the wind (gray contours) show two scales of vertical motion: 1) mesoscale terrain-tied ascent and descent in the southeasterly flow across Sardinia (nicely seen on 18 Nov) and 2) an eastward-propagating band of ascent associated with the aforementioned surface/850 hPa trough (front) and wind shift from southeast to west-southwest that approaches and crosses Sardinia between 1200-1800 UTC 18 Nov. Note also that the layer of conditional instability implied by the equivalent potential temperature cross section analysis is quite shallow and is mostly confined to below 800-750 hPa (http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/cross.html). Our interpretation of the Sardinia rain event was that it was associated with ordinary baroclinic cyclogenesis. The cyclone itself exhibited shallow warm-core characteristics below 800-750 hPa along and ahead of the of the surface/850 hPa trough. Because this warm-core structure remained shallow and widespread deep convection was not observed in the available satellite and radar imagery, we would argue that the event was not associated with a full tropical transition. The low-level warm-core structure likely developed ahead of the cyclone over the Mediterranean Sea and northern Africa in response to the passage of a mesoscale PV anomaly aloft that was rotating cyclonically around the southern side of the eastward-moving cutoff cyclone situated over the western Mediterranean Sea. We speculate that cooling aloft ahead of this mesoscale upper-level PV anomaly contributed to air mass destabilization ahead of the aforementioned 850/surface trough. We speculate further that surface sensible and latent heat fluxes over the warmer waters (~19-21 C) of the Mediterranean Sea between the coast of North Africa and Sardinia contributed to additional air mass destabilization. Vigorous mesoscale ascent in strong low-level southeasterly upslope flow over eastern Sardinia in the presence of standardized PW values between +2 and +3 sigma (http://www.atmos.albany.edu/student/kgriffin/maps/pw_stdanom/pw_stdanom_eur_loop.html) in a weakly conditionally unstable lower troposphere likely contributed significantly to the localized very heavy rains. On the larger scale, this case is illustrative of a baroclinic upper-level PV anomaly that fractures from the polar airstream in response to AWB and is subsequently influenced by additional AWB in the subtropical airstream in conjunction with midlatitude and subtropical trough interactions. The additional AWB in the subtropical airstream amplifies the downstream trough off the coast of northwest Africa and increases the environmental mean westerly flow in which the trough is embedded. Rich and interesting complexity is present throughout the event. Sorting it all out will be a challenge…. and an exciting research opportunity. Lance, Alicia, Ben, Julian, Philippe and Chip P.S. The remarkable and continuing uncertainty associated with the degree of phasing between disturbances in the northern and southern streams for the upcoming Tu/Wed pre-Tahnksgiving storm is nicely illustrated in Kyle Griffin's d(prog)/dt 500 hPa height/temperature and wind forecasts verifying between 72h and 180 h at 1200 UTC 27 Nov 2013 (http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/dprog/500vort_namer_dprog.html). Normalized spreads in the 72 h GEFS SLP forecasts verifying 1200 UTC 27 Nov are quite large (http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/map/gefs_1.png) while the 72 h GEFS normalized ensemble 500 hPa geopotential height spreads remain in the +2 to +3 standardized anomaly range (http://www.atmos.albany.edu/student/ppapin/lb13_img/mapdisco/20131122/images/map/gefs_2.gif) for the phasing of the troughs in the northern and southern streams on 1200 UTC 27 Nov. You can get any forecast answer you want in Alice's ensemble forecast restaurant…... 6. References: Mullen, Steven L., 1987: Transient Eddy Forcing of Blocking Flows. J. Atmos. Sci., 44, 3–22. Mullen, Steven L., David P. Baumhefner, 1988: Sensitivity of Numerical Simulations of Explosive Oceanic Cyclogenesis to Changes in Physical Parameterizations. Mon. Wea. Rev., 116, 2289–2329. By tradition, the last regularly scheduled Friday map discussion of the semester will be run by the students on 6 Dec 2013. Lance