Hurricane/Post Tropical Storm Sandy

(Click on images to enlarge)

Post Tropical Super Storm Sandy was an epic storm. It could very well be the worst storm that ever hit New Jersey, southern New England, Long Island, and New York City. It will take awhile to assess the total cost of the damage of the storm, but it will be in the billions. It cost dozens of lives and left millions in the dark. The transit system of New York City was crippled like never before.

Our forecast warning area was spared the worst of Sandy. That said, we certainly had issues to deal with, mainly with wind damage as well as tidal river flooding, one storm related fatality, but nothing like our neighbors to the south and east.

On October 22nd, the 18th Tropical System of the year, Sandy, formed south of Jamaica in the Caribbean Sea. It intensified fairly quickly into a Category 2 hurricane as it slammed into portions of Jamaica and Haiti. It then tracked north into the southeastern Cuba, after which it weakened to a low end Category 1 hurricane/high end tropical storm due to land interaction. It brushed the Bahamas and its outer bands of rain hit sections of the east coast of Florida, as well as North Carolina. An anomalous ridge of high pressure in the mid levels of the atmosphere forced the storm to track mainly due north very slowly, as its center remained off the southeast coast. Meanwhile, a high amplitude trough was located west of the entire Appalachian chain with surface highs anchored over Hudson Bay, as well as across Newfoundland.

The tropical system continued to track in a mainly northerly fashion October 26th through October 29th. At the same time, a significant upper level disturbance, including a very strong upper-level jetlet, took aim toward the mid-Atlantic states, poised to interact with Sandy. The high to the northeast strengthened, which ultimately prevented Sandy from heading out to sea. As the strong energy from the mid-Atlantic interacted with Sandy, it pulled Sandy back to the northwest, over the Gulf Stream, well off the Delmarva Pennisula. Sandy actually strengthened with core winds clocked to 90 mph before it made landfall near Atlantic City, New Jersey about 800 PM EDT on October 29th. At this point, it has lost some of its tropical characteristics but packed a stronger punch than a regular high end category 1 hurricane would have, due to the fact it had become fully absorbed into the mid-Atlantic system. A few hours prior to landfall, it had a central pressure of only 940 mb, one of the lowest on record. It became embedded within a very large mid latitude low pressure area which, in combination with the full moon occurring on the 29th, and the trajectory of the storm approaching from the east southeast, likely contributed to the record storm surge. Once it made landfall, post tropical system Sandy tracked west northwest across southern Pennsylvannia, then turned northward into western New York State on October 31st. It weakened considerably after landfall, and as dry air in the mid layers of the atmosphere worked into it.

While our region escaped the worst impacts from Sandy, record flooding from a tidal surge worked up the Hudson River, at Poughkeepsie. The reading at Poughkeepsie crested at 9.54 feet. The NYC Battery crested at 13.88 feet. Tidal flooding was noted all the way north to Albany. Minor flooding occurred on Lake George as water from a strong northerly fetch piled up into the village of Lake George. A few river gages in the Catskills came close to flooding but no points went above flood stage as rainfall amounts were not that heavy. In fact, much of region in the Hudson Valley saw rainfall amounts under an inch due to valley shadowing as easterly winds helped dry out the air, as it descended down off the Berkshires and Taconics. Also, drier air from the Canadian Maritimes was injected into the right side of Sandy, which helped mitigate rainfall amounts everywhere in our region which was on the right side of the storm track. The maximum rainfall amounts were a little over 3 inches in western Greene County.

The winds were forecast to be the worst impact from Sandy. While they did not quite reach the destructive levels as feared or forecast, we did have reported gusts 60-77 mph in our region, mainly across the higher elevations. A trained weather spotter located near Hancock, MA in the Berkshires, reported the highest wind gust reaching 77 mph. Most of our area had peak wind gusts in the 50-59 mph range. The official high wind gust at Albany International Airport was 49 mph, lower than the peak wind from Irene. Nevertheless, there were minor damage reports, mainly downed trees and wires, especially across the higher elevations and south of Albany, as well as isolated to scattered power outages. There was one storm-related fatality in our forecast area, a woman was killed in Kerhonkson, NY in Ulster County when a large section of roofing blew into her windshield. The worst of the effects of Sandy happened late Monday afternoon (Oct 29) through much of the night. By Tuesday, October 30th, as Sandy was well inland to our southwest, the wind initially decreased and so did the scattered rain bands. However, gusty southeast winds of 40 to 50 mph took place again midday Tuesday as good mixing developed after the sun came out, and temperatures remained very mild. These winds quickly diminished by dark.

The Operational ECWMF honed in on this storm from 10 days out, even before Sandy formed. That forecast model initially had the storm hitting Cape Cod on Day 10, a day too late. However, for subsequent runs after 00Z Tuesday October 23rd, the ECMWF had the storm tracking into the northern mid-Atlantic coast or New York City, with central pressures as low as 927 mb. Although the track wobbled a little during run to run, with one run considerably further east and others slightly to the south, making landfall at the mouth of the Delaware Bay, the operational ECMWF was superior to any other long-range forecast model. However, many other ECMWF ensemble packages, at least about a week out, still had the storm hooking to the east, so the operational European model was on the western side of its ensemble envelopes. The operational GFS initially tracked Sandy well out to sea, missing the east coast by a thousand miles! Many ensemble members in the early days (about a week before the event) did have the storm hooking back into Maine or Cape Cod. Several earlier runs of the GFS did produce an interesting trowal feature that would have brought a stripe of heavy rain to our region, but still kept Sandy well to our east. Within five days of the event, the GFS latched onto the event, with an increasing number of ensemble members hooking it westward into the region, mainly from Cape Cod northward. Slowly, the GFS honed in on taking Sandy into southern New Jersey. Overall, the GFS operational model was on the eastern side of its ensemble envelopes. Initially, the Canadian Global Model also had the storm hitting New Jersey, but later runs "lost" it to sea, before honing back in on it about five days out. The NAM ultimately did hone in on southern New Jersey within several days of the event. The United Kingdom Model (UK) initially had the storm tracking well out to sea, before ultimately tracking it into the Jersey shore in the shorter-term.

Six hourly balloon launches were mandated from 18Z Thursday October 25th through 06z Tuesday October 30th. This was the first time in history that ALL upper air stations in the United States were required to do six hourly balloon launches. Two staff members were deployed to the NY State Office of Emergency Management starting over the weekend of October 27 and 28th, 24/7 Saturday through Tuesday.

Forecast guidance

Above:  Loops of 500 hPa heights and vorticity from (left to right) the GFS, ECMWF, CMC and GFSEnsemble, initialized 0000 UTC 25 October.

Above:  Loops of MSLP and vorticity from (left to right) the GFS, ECMWF, CMC and GFSEnsemble, initialized 0000 UTC 25 October.

Above:  Loops of 850 hPa winds from (left to right) the GFS, ECMWF, CMC and GFSEnsemble, initialized 0000 UTC 25 October.

Above:  Loops of 925 hPa temperatures from (left to right) the GFS and ECMWF initialized 0000 UTC 25 October.



Above:  Sandy track forecasts from the National Hurricane Center displayed on AWIPS (left and middle) at 1500 UTC 29 October and 2100 UTC 29 October, and from 2100 UTC October 29 displayed in Hurrivac with the cone of uncertainty (right).



Above:  Loop of forecasted track of Sandy from the National Hurricane Center beginning 2100 UTC 29 October (left) and loop of the GFS and ECMWF MSLP beginning 0000 UTC 22 October valid 0000 UTC 30 October (right).


Above:  Loop of MSLP, 700 hPa heights and 850-700 hPa 2-D frontogenesis from the 1800 UTC 29 October GFS.

Above:  Loop of MSLP from the 1800 UTC GEFS.




Above:  Loops of 1800 UTC 28 October GEFS 850 hPa winds and anomalies (left) and 1500 UTC 28 October SREF 850 winds and anomalies (right).



Above:  Loops of 850 hPa winds from the 1800 UTC 29 October GFS (left) and NAM (right).



Above:  Loops of 925 hPa winds from the 1800 UTC 29 October GFS (left) and NAM (right).



Above:  Loops of cross sections of frontogenesis between the southern New England coast to Lake Ontario from the 1200 UTC October 29 NAM (left) and 1800 UTC 29 October GFS (right). 





Above:  Graphic of the wind impacts of Sandy.



Above:  Precipitation totals from 3 different perspectives.



Above from left to right:  Dual Pol precipitation estimates, legacy Doppler radar precipitation estimates and Vermont and northern NY precipitation totals.




Above:  Displays of river levels from assorted points in the forecast area.




Above:  Displays of river levels from assorted points in the forecast area and graphic of the flood warning for Poughkeepsie.


Photos of flooding in Poughkeepsie



Links to wind, rain and damage reports


ASOS reports


Local Storm Report


Assorted rain reports


Assorted wind reports


Post storm hurricane report


Power Outages