8-9 February Historic Snowstorm


After a fairly quiet January, the models indicated some sort of fairly robust system would impact our region with snow on Friday Feb 7th well over a week out. Early runs of the GFS indicated this would more of a clipper system with light amounts of snow, possibly mixing with rain south. The ECMWF for several days indicated a more significant storm. It had this storm receiving energy and moisture from the southern stream, further south and with more QPF. Initially the NAM did not have a clue on the evolution and track of the storm, bringing an open wave in fast Thursday overnight and having it rapidly exit to the east by Friday.


Once again in the longer term, while not perfect, the operational ECMWF outperformed all the other models. It was the first model to suggest some sort of transition or phasing of two systems (the northern one) and another separate one in the southern stream, to develop into one potent coastal storm somewhere east of Cap Cod.


The NAM finally caught onto the coastal low and then ramped up the QPF as high as 2-3 inches across our entire CWA which turned out to be way too high. The operational GFS probably had the best short term QPF forecast, although with this case it was a little low across the northwest and southeast portions of our County Warning Area, but very close to reality in the Greater Capital region. It appeared to be right for the wrong reasons, as it too seemed to miss the down sloping or shadowing effect from the Southern Greens and Berkshires.


None of the models was perfect. What ended up happening is that a potent system that came from the Pacific, producing blizzard conditions across much of the west, maintained its strength as it worked into western New York and even northern Vermont on Friday morning. One interesting band of snow extended into Maine with this system.


Meanwhile, a weak system in the southern stream ramped up, including lots of convection on Friday as well. This system deepened off of Cape Hatteras Friday morning. There was some concern that the models did not analyze the exact location of the initial low pressure as it was perhaps 25-50 miles too far east compared to the actual low. However, as is often the case, the models quickly corrected this feature.


The big challenges of this storm were when the northern branch would phase with the southern storm. Some models did this too quickly. What ultimately resulted was the northern stream system remained positively tilted which was the impetus that kicked the rapidly deepening low offshore. The two systems did phase early Friday when the storm was east of Cape Cod. The coastal low had deepened to 982 hPa even before the phasing took place, then slightly filled to 985 hPa according to the MSAS MSLP analysis. The Rapid Refresh (formerly RUC) had the storm deepening Friday morning. After the phasing, the combined storm did drop to about 975 hPa but by this point, only our eastern zones were underneath its grip.


The northern stream system was responsible for producing snowfall the majority of snowfall in our northwestern areas. It was critical that a deformation band produced enough snowfall Friday morning to put some of the zones north and west of the Capital District into the 9" criteria for 24 hours. Some examples were Schoharie and Hamilton Counties. The complex interaction between the two systems produced enough ascent (along with isentropic lift) to produce snow in the Capital region in the 6 to 12 inch range...while the actual cold conveyor belt tied in with the coastal low pounded our southern and eastern areas with a foot or more of snow. Litchfield was the winner of the most snow...with 33 inches falling in New Hartford and most of the county reported around to two feet of snow!


Despite the precipitation being all snow in our entire CWA, there were significant challenges forecasting this storm. Besides trying to determine when the two systems would phase, other challenges with this storm include where the mesocale snow bands would set up, the ultimate QPF, the snow to liquid ratio and in the Hudson valley, there was a period of precipitation shadowing with the northern stream storm. Finally, the radar appear to be running "hot" at times, indicating 20 DBZ values over the office, when in fact, only light snow or even flurries were falling, although the flakes were bigger at this point (melting layer or bright band effects possibly).


Model soundings from the NAM and the GFS were not totally representative of what happened. These suggested snow to liquid ratios would be on the order of 12-15:1. One forecaster tried utilizing the Caribou Snow tool during the height of the storm and came up with nuhPaers too high. As it turned during much of the time, the best ascent in the column (at least in the Albany area) fell in the area considerably warmer than -14 C (where the best snow growth area is realized). On the other hand, the Adirondacks had much fluffier snow in the order of a snow to liquid ratio of 15:1


During the afternoon into the early evening hours, VAD winds of the ALY RAOB indicated a deep easterly wind flow from about 1-4 thousand feet. (the surface wind at the surface was more from northeasterly direction, at least at the Albany International Airport). The easterly flow was enough to dry the lower levels of the column enough to prevent a lot of snow from accumulating in the Hudson Valley around the Capital District and point north. The shadowing was not as much of an issue further south in the Hudson valley. As a result several hours of accumulating snow was lost. Later in the evening, as the overall flow became more northeasterly and the lower levels of the column finally saturated, is when the best snowfall took place in the Capital region. The flakes however were small during this time, indicating that the ascent was not in the optimal snow growth area. Later that evening, two bands formed, one to the west and eventually a bigger one to the east of the Capital District. Snowfall rates underneath this band, were estimated to be 1-2 inches per hour. The lighter snow which fell over the most of the Capital region nevertheless produced snowfall rates of between half an inch and one inch per hour. The cold conveyor band, likely producing snowfall rates 3-4 inches per hour, remained mainly over Litchfield and eastern Berkshire County. The storm was too far east of the 40/70 bench mark to produce any real significant Mohawk Hudson convergence. Due to being this far east, the surface winds to the west of the Hudson valley were not all that strong and the inversion of our region was mixed out.


2. What was learned from this event: Liquid to snow ratios do matter as well as QPF. The NAM once again proved it has serious issues forecasting QPF. It might be better to look at the overall synoptic pattern and try to identify the mesoscale and orographical features that could enhanced QPF.


The models did not do well in forecasting snow ratios. The Real-time Snow Ratio Forecast Page indicated there was greater than a 67 percent probability the snow-liquid ratios would be at least 15:1 at Albany during the height of the storm, when in fact this ended up being way too high. This was yet another factor which contributed to low snowfall totals in the Greater Capital District.


Graphics - 500 hPa forecasts - note all the sources of guidance show the phasing of he northern and southern stream systems as it is tracking off the New England coast, not an optimal scenario for big snows in eastern NY.



Above:  Loops of GEFS (left) and SREF (right) 500 hPa heights and anomalies.



Above:  Loops of ECMWF 500 hPa heights valid at 1200 UTC 9 February (left) and 0000 UTC 10 February (right).


Graphics - Mean Sea Level Pressure forecasts - The models took considerable time to resolve the surface low tracking off Cape Cod but you can see the subtle differences in predicted storm track and intensity, differences that could have a big impact in how far north and west the heaviest snow would fall.


Above:  Loops of GEFS (left) and SREF (right) MSLP.



Above:  Loops of ECMWF MSLP valid at 1200 UTC 9 February (left) and 0000 UTC 10 February (right).


Graphics - 850 hPa winds and anomalies - Note again that the models did not resolve the strong easterly wind anomalies until about 2-3 days prior to onset and did not resolve the maximum magnitude until 12-24 hours prior to onset.  The orientation and westward extent of the maximum anomalies suggested New England would receive the heaviest snow amounts but even eastern NY was within the 4 SD contour so again, the westward extent of the heaviest snow was in question.



Above:  Loops of GEFS (left) and SREF (right) 850 hPa winds and anomalies.


Graphics - probabilities for 1 inch of liquid equivalent precipitation in 24 hours and 2 inches in 36 hours - There were disagreements in the northwestward extent of the 1 inch and 2 inch probabilities, and changed from run to run.  Because of the averaging done in ensembles, the heaviest precipitation is usually a little north and west of the where the ensemble guidance depicts the probabilities.



Above:  Loops of GEFS (left) and SREF (right) probabilities for 1 inch of liquid equivalent precipitation in 24 hours.



Above:  Loops of GEFS (left) and SREF (right) probabilities for 2 inches of liquid equivalent precipitation in 36 hours.


Graphics - Plume diagrams



Above:  Loops of plume diagrams for Albany, NY from the GEFS (left) and SREF (right).




Above:  Loops of plume diagrams for Bradley, CT from the GEFS (left) and SREF (right).




Above:  Loops of plume diagrams for Monticello, NY from the GEFS (left) and SREF (right).


Graphics - Model data from AWIPS



Above:  ECMWF and GFS initialized 1200 UTC 3 February (left), GFSEnsemble mean MSLP and members initialized 1200 UTC 3 February (center), and the SREF mean MSLP initialilzed 1500 UTC 7 February.



Above:  Four-panel displays of the GFS, NAM (12 km and 80 km) and ECMWF initialized 1200 UTC 7 February of MSLP (left), 850-700 frontogenesis (left

center), QPF (right center), and pure model snowfall (right).



Above:  Four-panel displays of the GFS, NAM, ECMWF and GFSEnsemble and members initialized 1200 UTC 8 February of MSLP



Above left:  Four-panel display of the GFS, NAM, ECMWF frontogenesis and temperatures initialized 1200 UTC 8 February.  Above center:  Four-panel display of the GFS, NAM, and ECMWF QPF and GFSEnsemble probability for 1 inch of liquid equivalent precipitation in 24 hours initialized 1200 UTC 8 February.  Above right:  Four-panel display of the GFS, NAM, and ECMWF QPF and SREF probability for 8 inches of snow in 12 hours initialized 1200 UTC 8 February.



Above left:  Time sections from the NAM (left) and GFS (right) for Albany, NY.  Notice the core of the maximum vertical motion is through the maximum dendritic zone but is above 700 hPa. 



Above left:  Time sections from the NAM (left) and GFS (right) for northwestern CT.  Notice the core of the maximum vertical motion is through the maximum dendritic zone and right around 700 hPa. 


Loops of data



Above:  Loops of water vapor satellite imagery.



Above:  Loop of Infrared satellite imagery.



Above:  Loop of visible satellite imagery.



Above:  Loops of 2 different types of MSAS surface analyses:  MSAS with 3 hour pressure tendency (left) and MSAS with surface observations and pressure tendency (right).



Above:  Loop of KENX Velocity Azimuth Display Wind Profile.



Above:  Regional radar mosaics.



Above:  Regional mosaic of radar imagery courtesy of the College of Dupage.



Above:  Radar estimated snowfall from KENX.



Above:  Plots of snowfall across the northeastern U.S based on all snowfall observations.



Above:  Plot of snowfall from the National Climatic Data Center and NESIS rank based on CO-OP snow observations only.