March 12-13 2014 - Snow and Mixed Precipitation Event

On the morning of March 12th, low pressure was over West Virginia. A cold front was pushing south of Interstate 90, gradually bleeding another relentless round of arctic air south from Canada. The air was initially not all that cold, but precipitation was either rain (from Albany south) or snow (north of Albany) as the column was nearly isothermal. As the low intensified while tracking across southern PA, cold air continued to wedge into the lower levels of the atmosphere. At the same time, southerly winds aloft began driving a warm nose from the southeast, into the Capital Region and eventually even points to the north of Albany, changing snow in those places to freezing rain. The only exception was in the Adirondacks where a significant snowstorm continued for the duration of the event. Rain changed to freezing rain in the Capital Region around 200 PM. At the same time however, temperatures were in the 40s with rain across much of the Berkshires, and even into the southern Green Mountains. Temperatures in the 40s were still common south toward Poughkeepsie was well. North of Albany, temperatures were down into the 20s at this point.

Initially, at least right in the Capital District, icing was slow to get started due to the stronger March sun angle (even though it was obscured by clouds). Also, temperatures had risen into the 50s the previous day, and remained above freezing throughout the previous night. However, as temperatures dropped into the upper 20s and we began to lose the sun angle, as icing did commence by late afternoon. What was surprising is that icing up to a quarter inch was reported as far north as Gansevoort. Most the of the higher elevations in and around the Capital District had around a quarter inch of ice, some locally higher amounts closer to half an inch. In the lower elevations of the Capital District and points to its south, around a tenth of ice was reported.

By late afternoon, the storm tracked just north of Philadelphia, and eventually headed toward New York City by evening. At this point, the warm nose was gradually shunted, the mid-levels of the column began to cool, and precipitation transitioned into sleet Albany southward, and went back to snow (or a snow sleet mix) further north. Cold air rapidly invaded all areas to the east and eventually well south of Albany through the evening. Temperatures fell into and in some cases, through the 20s. A coating of sleet (at in the Capital region) probably gave a little more traction on surfaces that otherwise would have been pure glaze.

By midnight, the storm was near Cape Cod and all precipitation switched to snow. While most of it was light, an interesting narrow band formed from southern Saratoga southwestward into portions of the Greater Capital District. This band ended up right over the airport, producing 3 inch snowfall amounts in one hour overnight! Other areas reported much lighter snowfall amounts, generally well under an inch per hour. The on-going snow was partially due to the fact the mid-level trough lagged behind the storm, (never capturing it, at least until it was well to our east). This upper level trough added additional ascent until it moved to our east later on Thursday.

Another larger, but less intense band formed to the east of the Capital District toward daybreak. This band might have been tied in with Mohawk Hudson Convergence (the usual suspects, a northwest wind at KRME, northeast or east wind at KGFL and North wind at KALB) all of which was happening. In addition, since it was not confined to the Capital District, an upslope component was likely tied into it as well. It produced snowfall rates of no more than an inch per hour (if that). The first band was not predicted by any models (including the HRRR) but apparently the HRRR at least hinted to the second bigger band for quite some time.

Wind played a somewhat a role in the storm. For a while in the evening, a north wind gusted up to 35 mph even right in the Capital District. Other places occasionally reported similar gusts from a north (valley locations) or a northeast direction (higher terrain). There appeared to be a little lull in the wind, at least locally, around midnight, but once the wind came around to the northwest they increased up to gusts in excess of 40 mph at times. The wind produced blowing and even drifting snow from the CD northward, and added misery to an already anomalously cold air mass for the time of year.

The storm finally ended by midday Thursday (March 13th) and more than a foot of snow fell across much of our Adirondack Park. A spotter near Holcombville (Gore Mountain) reported 23 inches! Nearly a foot of snow also fell in extreme northern Windham and Bennington counties.

A sharp line of delineation took place further south. The northern rim of the Mohawk Valley...much of the remainder of Southern Vermont generally reported 4 to 8 inches. Further south, including the southern portions of the Mohawk Valley east to most of the Capital District Berkshires, extreme southern Vermont, Catkskills and the Mid-Hudson Valley)...snowfall accumulations generally decreased to the 1-4 inch range (except that narrow band discussed earlier that officially brought the total snowfall at Albany International Airport to 5.5 inches).

Model data turned out to be exasperating up to the day of the event. Initially the European Model (ECMWF) had a blockbuster storm for most if not all our region, while the GFS had a much weaker wave gliding harmlessly out to sea. Neither solution was correct. A run or two of the ECMWF did track the storm much further inland (but always south of the CD) which appeared a little unusual (but turned out to be correct). However, the ECMWF model reverted back to tracking the low along the classic benchmark, offshore to central Long Island and the tip of Cape Cod. It offered this solution for three consecutive runs. This increased confidence that this would be a mainly snow event and lots of it for at the CD and points northward.

The models had a lot of trouble with the track and the thermal profile, even a day before the event. The 06Z GFS/NAM runs on the 11th were vastly different. The GFS was much warmer (having finally resolving the storm), implying a lot of rain from the CD southward, while the NAM was much colder, still suggesting all snow from the CD northward. The ECMWF began trending a little northward, but still was suggesting a potential range of precipitation types that might dominate.

The 12Z NAM then came in with a difference of +11C at the H850 during the peak of the storm over the CD! Now it implied all rain with the storm tracking very close to the CD (turned out to be a little too far north). The 12z GFS (and it ensemble members) were fairly consistent from the previous run. The 12z ECMWF trended warmer. The Canadian Model (CMC) was as usual, all over the place with some runs having it go out to sea, some runs burying us, and a few actually tracking the low to the north of Albany. The 18Z NAM on the 11th came in colder again, (not as cold as before) but definitely decisively colder than the 12Z run.

It was not until the day of the storm (March 13th) that the RUC finally caught on closer to reality, however it did not pick up on the skinny meso-scale band that formed later at night. The track of the 700 Mb low was consistently indicated well north of the Capital District in most sources of guidance prior to the event and turned out to be a good indicator of where the line of demarcation took place between the heavier and much less heavier snowfall amounts.

Weather information we provided to the user community consistently stated that forecasting this was lower confidence than usual since a 50 mile change in the track would ultimately make a HUGE difference in snowfall amounts which is exactly what happened. The heavy snowfall amounts (even with near blizzard conditions) were confined to a place that could better handle this amount of snow.

Model MSLP forecasts



Above:  MSLP forecasts from the 1200 UTC 9 March (left) and 1200 UTC 11 March (right) GFS (upper left), NAM (upper right), GFSEnsemble mean and spread (lower left) and SREF mean (lower right).

Above:  MSLP forecasts from the 1200 UTC 9 March (left) and 1200 UTC 11 March (right) ECMWF.

Ensemble Guidance - notice the abrupt shift northward of the snowfall predictions as the onset of the storm approached.

Above:  Ensemble probabilities from 0900 UTC March (left) and 0900 UTC 11 March (right) GFSEnsemble probability for 1 inch liquid equivalent precipitation in 24 hours (upper left), SREF probability for 1 inch liquid equivalent precipitation in 24 hours (upper right), SREF probability mean 12 hour snowfall (lower left) and SREF probability for measurable snow (lower right).

Forecast soundings - notice the column gets above freezing at KALB and KPOU but stays below freezing at KGFL.

Above:  Loops of forecast soundings from the 0000 UTC 13 March ECMEF, NAM and GFS from Albany, NY (left), Glens Falls, NY (center) and Poughkeepsie, NY (right).

Plumes - Notice the clustering of liquid equivalent precipitation increased rapidly just prior to the onset of the storm and the SREF was

indicating rain further south than the GEFS.

Above:  Loops of 1200 UTC 8 March GFS Ensemble plumes from Albany, NY (left), Watertown, NY (center left), Burlington, VT (center right) and Monticello, NY (right).

Above:  Loops of 0900 UTC 9 March SREF plumes from Albany, NY (left), Watertown, NY (center left), Burlington, VT (center right) and Monticello, NY (right).

Above:  Loops of snowfall forecasts (left) and ice forecasts (right) from the Weather Prediction Center between 1843 UTC 10 March through 0612 UTC 12 March.  Note the axis of the heaviest snowfall is consistently predicted over the northern half of the NWS Albany forecast area but the southern gradient tightens closer to the onset of the storm as the varying freezing rain forecasts shift through southern NY and New England.

Observations and Model Initializations at the peak of the storm - Note the 850 hPa U wind anomalies were over far northern NY.  Also note the subtle dominance of the 850 hpa V wind anomalies in the SREF which implies stronger warm advection than what the GEFS suggests.

Above:  Analyses of 500 hPa heights and anomalies from the 1200 UTC 13 March GFSEnsemble (left) and 0900 UTC 13 March SREF (right).


Above:  Analyses of MSLP and anomalies from the 0000 UTC 13 March GFSEnsemble (left) and 2100 UTC 12 March SREF (right).

Above:  Analyses of MSLP and anomalies from the 0000 UTC 13 March GFSEnsemble (left) and 2100 UTC 12 March SREF (right).

Above:  Soundings from 0000 UTC 13 March from Albany, NY (left), Buffalo, NY (center) and Upton, NY (right).

Above:  Loop of water vapor satellite imagery through the storm.

Above:  Loops of visible satellite imagery through the storm.

Above:  Loop of radar reflectivity during the end of the storm showing a highly localized mesoscale snow band, supported by Hudson/Mohawk convergence.

Above:  GIS analysis of observed snowfall.