Historic End of February 2010 Storm:  Stage 2 (Click on images for larger display)


This was an extremely challenging and long lived event. The longest Winter Storm event since the December 11-14, 1992 Winter Storm. This storm also involved multiple hazards including heavy snow, high winds and heavy rainfall. The highest snowfall totals occurred across the southern Montgomery...Schoharie...western Schenectady...western Albany...western Greene and western Ulster County where snowfall accumulations of 2 to 4 feet were common (Hunter Mountain 48"). The most unusual feature of this storm was the track of the surface low moving westward across southern New England as well as the oscillation of the 500 hPa low and both of these features impacting our region for 72 hours.  As stated earlier, this storm was being advertised in a range of terms from “Significant storm”, to “Big Storm”, to “February Fury”, to “Snowicane” by different sources of weather information.  This caused a debate within the meteorological profession on how to communicate significant hazards to the user community.  Some commentary can be found here.


An analysis of 1200 UTC 02/25/10 model guidance revealed that the NAM had the best position and strength of surface low through 1800 UTC 02/26/10, while the GGEM had the best position and strength of the surface low from 0000 UTC 02/27/10 through 0000 UTC 02/28/10. The surface low which was initially still connected to the inverted trough that was responsible for the storm earlier in the week, tracked from east of Delmarva on the morning of the 25th north to south of Cape Cod on the evening of the 25th. This system then moved westward to near New York City on the morning of the 26th and stalled over extreme southeast New York and northern New Jersey during the day on Friday before lifting north to near Albany on Saturday. The lowest pressure of 972 mb was observed near KBDR at 0600 UTC on 02/26/10.


An analysis of the 500 hPa heights indicated that all of the model guidance (NAM, GFS, ECMWF and GGEM) initialized correctly with the lowest heights across southeast Ohio at 1200 UTC 02/25/10, but once again the NAM was superior in the placement of the lowest heights as well as the configuration of the mid level trough axis until 1200 UTC 02/26/10. From 1200 UTC 02/26/10 forward the models had difficulty with the location of the 500 hPa closed low and associated trough and by 0000 UTC 02/28/10 only the GGEM had this feature still within New York State. The closed 500 hPa low moved from southeast Ohio on the morning of the 25th to near KDCA that evening and well east of the New Jersey coast on the morning of the 26th. This feature then lifted north to central Long Island on the evening of the 26th and Albany on the morning of the 27th before moving westward to near Rochester that evening.


QPF fields from the NAM were overall superior during this event to those of the GFS and ECMWF and the forecasters working this event realized early on that the NAM was handling the surface and upper level features as well as QPF better. Thus, the gridded forecasts were heavily based on the NAM fields. The NAM QPF fields for the first 24 hours ending at 1200 UTC 02/26/10 did an excellent job of showing the enhanced upslope precipitation across Eastern Catskills, Mid Hudson Valley, Litchfield Hills and southern Berkshires and even highlighted the rain shadow west of the Berkshires and Southern Green Mountains. The GFS actually did better on the axis of heaviest precipitation on the second day ending at 1200 UTC 02/27/10, but the NAM was better on amounts except too little QPF in the valleys and too much across the Adirondacks. The heaviest precipitation on the second day fell from the south side of the Mohawk Valley and Northern and Central Catskills east to Southern Vermont and the Northern Berkshires. The model QPF for the third day ending at 1200 UTC 02/28/10 was poor from all of the models and this was likely due to the poor placement of the 500 hPa low and associated trough. The QPF amounts from the GFS were much too high across the Berkshires and southern Vermont and too light across the Adirondacks. The NAM picked up the additional QPF well across the Eastern Catskills, but also was similar to the GFS with respect to the Adirondacks, Southern Vermont and the Berkshires.


The NAM was far superior to the GFS for snowfall amounts and for the snowfall totals ending at 0000 UTC 02/28/10 it pinpointed the 30-40 inch amounts that fell from the south side of the Mohawk Valley through the Eastern Catskills and was fairly accurate elsewhere except for northern and central sections of the Hudson Valley where amounts were generally 2 to 4 inches too high.


Highlights of the Event:


1. States of Emergency were declared for Greene County as well as for parts of Albany, Schoharie and Dutchess Counties (this needs to be verified)


2. Numerous roof collapases occurred in parts of Albany, Schoharie, Montgomery, Greene and Ulster Counties. Many of these were abandoned buildings or barns. A couple was trapped in their house in Schoharie County for several hours and 4 cows were killed when a barn collapsed in western Albany County (this also needs to be confirmed)


3. Southwest Dutchess and Southeast Ulster County presented the biggest forecast challenge as a few hundred feet in elevation separated a few inches from 12 to 24 inches of snow accumulation. This is beyond the capability of our current models. The NAM at 1200 UTC 02/25/10 had a small sliver of 15+ inch amounts in 24 hours across the extreme southwest corner of Dutchess County but also had some higher amounts elsewhere in the county that did not verify. Rich Bodin, SKYWARN EC for Dutchess County, put it best when he estimated nearly 2 feet of snow at his house at a higher elevation in Wappingers Falls, but down the road from his house in the valley there was virtually no snow. While we ended up verifying Eastern Ulster County and taking a hit in Western Dutchess County, making a different decision on issuing warnings would have been difficult given our current zone configuration and time when much of the snow fell from the evening of the 25th into the early morning hours of the 26th.


One question that we need to answer: Once we get reports of warning level snows in a county, and heavy precipitation is expected to end (light, non-warning additional snows expected), should we issue a Winter Storm Warning, knowing it has little chance to verify with additional 7", yet the impact was severe. Do decision makers need certain headlines to justify certain courses of action for clean-up and/or rescue operations? Dutchess was under a Winter Weather Advisory, and could have been upgraded once we received observations of extreme snow amounts, which were severely lacking until 12"+ amounts started being reported in limited areas of the county.


Another question that we need to answer: How long after the heavy precipitation ends should we continue the Winter Storm Warning headlines? Neighboring offices dropped their headlines long before we did. Granted, we did have snows that just reached advisory level in the 24-48 hours following the heaviest precipitation in much of our forecast area. There was the threat of roof collapses and some travel impacts due to the additional advisory level snows. Also, we do not have to verify continued Warnings once they are verified, and most of the Warnings had been already verified. So again, how long do we hold onto Warning and Advisory headlines?


4. The High Wind Warning was probably issued for too large of an area (still waiting on feedback from power companies). Parts of Southern Vermont were hit very hard mainly across the east facing slopes of the Southern Green Mountains and in channeled flow through the mountains. The owner of the Prospect Mountain Cross Country Ski Center reported the strongest winds that he has ever witnessed in the 23 years that he has been at that location. Numerous trees, tree limbs and power lines were down in his area. The winds on Bromley Mountain were so strong that the lifts were not running (gusts in excess of 50 mph is their criteria).


5. Some ponding of water occurred with this storm as storm drains and culverts were blocked from the heavy snow which fell during the previous storm. A Flood Advisory was issued for the areas from the Hudson Valley east through Western New England at 510 PM Thursday Evening until 500 AM Friday morning. While Flood Warnings were issued Thursday evening and verified at Brookfield (moderate flooding) and Stevenson, no small stream or river flooding occurred in Litchfield County.


6. Given the complexity of the storm the forecast shifts did a good job of separating the area of mostly rain during the first half of the event from the area of mostly snow which eventually made it to the valley floor by Friday afternoon. Also, the areas which received the greatest snowfall were included in the first warning which ended up having the greatest lead time.



Above:  Loop of 500 hPa heights and vorticity from the GFS, ECMWF, GGEM and GFSEnsemble.  Note the differences in evolution toward the end of the loop, contributing to uncertainty in the storm track and ultimately the types of impact for different regions of the northeastern U.S.



Above:  Loop of MSLP from the GFS, ECMWF, GGEM, and GFSEnsemble.  Note again the inconsistencies toward the end of the loop leading to considerable uncertainty.


Above:  Loop of 500 hPa heights, spread and anomalies (color shaded) valid at 00Z 26 February from the GEFS, showing the considerable shifts in the forecasted position of the upper level system with each run, again, contributing to considerable uncertainty.


Above:  Loop of 850 hPa temperatures, heights and anomalies (color shaded), valid at 06Z 26 February from the GEFS, showing the significant differences in 850 hPa temperatures with each run.  There was a gradual trend for 850 hPa temperatures to be below freezing in the Catskills and southern NY, with a very sharp thermal gradient, meaning a very sharp rain/snow line in a very unusual configuration, oriented from NNW to SSE.




Above:  Loops of 850 hPa wind barbs and anomalies (color shaded), initialized from the GEFS at 00Z 21 February and 12Z 23 February.  Note the earlier GEFS run did not show quite as strong U wind anomalies, and did not resolve the 2 separate events.  The later run showed much stronger U wind anomalies, and showed 2 separate centers of anomalies distinguishing the 2 events.



Above:  Loops of 925 hPa temperatures, heights and anomalies valid at 00Z 24 February and 06Z 26 February, showing the first event was generally all snow across the region, and evaluating the rain/snow line was much more difficult for the second event.  



Above:  Loops of GEFS 24 hour probability of 1.00” liquid equivalent valid at 18Z 24 February and 12Z 26 February, showing the high level of uncertainty and gradual trends in each storm for the heavier precipitation to occur over a progressively larger area, expanding north and west.




Above:  Loops of SREF 24 hour probability of 1.00” liquid equivalent valid at 18Z 24 February and 12Z 26 February, showing the high level of uncertainty and gradual trends in each storm for the heavier precipitation to occur over a progressively larger area, expanding north and west.



Above:  Wind barbs and anomalies (color shaded) at 250 hPa.  The left 2 panels are forecasts from the 00Z 23 February NAM and GFS, valid at 12Z 26 February.  The right 2 panels are initialized from the GEFS and SREF at 12Z 26 February.  The 250 hPa wind anomalies supported a long duration event, as the storm was very much cut off from the steering flow.



Above:  Visible satellite image at 1701Z 26 February.  Note the areas of overcast and clearing within the circulation.  The storm was beginning to weaken at this time.



Above:  Water vapor satellite loop from @2300Z 25 February to @0800Z 26 February with lightning overlay.  This was the period of maximum strengthening of the storm.  Notice how it winds up into a classic spiral configuration.  Also note the dry slot rapidly tracking westward through New England at the end of the loop.



Above:  Loop of KENX base reflectivity at 0.5° with LAPS surface temperature overlayed from 1700Z 25 February to 0700Z 26 February.  Note the extremely tight thermal gradient through eastern NY, in some cases resulting in some counties experiencing rain in some parts to feet of snow in other parts of the counties.



Above:  Loop of regional radar from 0800Z 25 February to 1000Z 26 February, showing the heavy precipitation tracking through the region for >24 hours.  This precipitation loop preceded and is separate from the 3rd stage of the storm, the development of upper deformation precipitation, that added another 3-7” of snow during the next 24-36 hours over most of the region. 



Above:  Preliminary NESIS ranking of category 3, included all three stages of the precipitation from the storm.  It could be argued that the first 2 stages were separate events, and stages 2 and 3 were associated with the same upper level system, but for the purposes of the NESIS scale, all three stages were considered one long duration event.



Any comments, questions or suggestions, please mail to:Neil.Stuart@noaa.gov