Hurricane Sandy Summary
For many years, the New York City area was said to have been overdue for an extreme hurricane; there have been many tropical cyclones over the last few decades, some with significant impacts, although most were weakening extratropical-transitioning cyclones once affecting the region and generally had more impacts in terms of flooding, or kept the worst impacts east of NYC. The last hurricane to hit the eastern United States was hurricane Irene in 2011; Irene made landfall in NYC as a tropical storm, and resulted in significant impact especially from flooding across the region, most notably in interior New England with devastating floods, but did not produce the worst possible impacts for coastal locations. Hurricane Sandy, however, took a different approach when affecting the region compared to typical East Coast hurricanes, with stronger intensity, longer duration, and extreme damage along the coast that have secured its place as one of the worst storms on record to affect both the NYC area, and with over $60 billion in damage, one of the most damaging storms in the United States as well, with Sandy the second costliest US hurricane in recorded history, only behind Hurricane Katrina.
Hurricane Sandy Archive
October 25 – Historic Storm Early Next Week
October 26 – Hurricane Sandy Approaches
October 27 – Historic Impact from Hurricane Sandy
October 28 – Final Hurricane Sandy Forecast
Hurricane Sandy Updates
Hurricane Sandy History:
Hurricane Sandy was a late season tropical cyclone that developed out of a tropical wave in the Caribbean Sea, following days of signals for a tropical cyclone to develop in the Caribbean sea around that time frame. The wave gradually continued to become better organized, and on October 22 was designated as Tropical Depression 18. TD-18 was initially nearly stationary, slowly drifting to the SSW, and intensified enough to become a tropical storm later that day, when it was named Sandy, the eighteenth named storm of the 2012 hurricane season. Even before Sandy’s formation, the model guidance had already suggested that Sandy would directly affect the East Coast afterwards with extreme impacts, a rather impressive signal nearly ten days away from impact.
By October 23, Sandy slowly intensified as a tropical storm while moving steadily to the NNE as it approached Jamaica and Cuba. Early on October 24, Sandy began to intensify more steadily as it approached its first landfall; by 3 PM, it made landfall in southeastern Jamaica as a category 1 hurricane with 80 mph sustained winds. Once re-emerging over water, Sandy rapidly intensified in the narrow stretch of water between Jamaica and Cuba, making its second landfall only hours later as a 115 mph category 3 hurricane. Initially, Sandy was reported to have been a category 2 hurricane at landfall, but was upgraded to a major hurricane following post-storm analysis in early 2013. Following the Cuba landfall on the 25th, Sandy’s structure became more disorganized, somewhat resembling a subtropical cyclone in appearance, while increasing in size and slowly tracking towards the northwest through the Bahamas. Meanwhile, Sandy was affected by stronger wind shear and slowly weakened into a category 1 hurricane; early on the 27th, Sandy briefly weakened into a strong tropical storm, although it quickly re-intensified into a hurricane while maintaining sustained winds of 75 mph but with a steadily decreasing minimum pressure, resulting in an increasingly large size.
Throughout the 27th and 28th, Sandy continued to slowly track to the northeast parallel to the East Coast but well offshore. Sandy was a weak category 1 hurricane during this time frame struggling with wind shear and dry air but still had a low minimum pressure, around 960-950 millibars, in relative to its sustained winds of 75 mph, along with an abnormally large size; during this time frame, Sandy had one of the highest integrated kinetic energy (IKE) values of any tropical cyclone in the Atlantic Ocean in recorded history, even more than other devastating hurricane such as Katrina of 2005 and Ike of 2008. During these two days, a highly anomalous set-up developed which would later force Sandy to take a very unusual track for a tropical cyclone. Normally, tropical cyclones especially in late October continue to track northeast and well offshore, but Sandy was prevented from doing so; anomalous blocking developed near Greenland, later shifting towards the Davis Strait and Newfoundland, while an upper level low remained stationary south of this block over the northern Atlantic. Ridging was present east of Sandy, which would later connect with the blocking to the north, acting to prevent Sandy from escaping east. A trough dropping into the central and eastern US at the same time ended up capturing Sandy, forcing it to retrograde sharply to the west, aiming directly at the New Jersey coast while intensifying and transitioning into a massive hybrid cyclone of tropical and non-tropical characteristics.
As Sandy curved to the northwest on October 29, with the aid of baroclinic forcing, lower wind shear, and the warm gulf stream waters, it quickly intensified off the Northeast US coast into a category 2 hurricane with sustained winds of 100 mph and an unusually low minimum pressure of 940 millibars. At this point, with a wind field of over 1,000 miles in diameter, Sandy became the largest Atlantic hurricane on record. Sandy continued to lose tropical characteristics as it accelerated to the WNW, and was declared extratropical during the evening while 20 miles south of Atlantic City, just off the New Jersey coast; an hour afterwards, it made landfall at 8 PM just southwest of Atlantic City with sustained winds of 80 mph and a minimum pressure of 946 mb, a record low for Atlantic City. Hurricane Sandy officially lost tropical characteristics just an hour before landfall; the last hurricane to make landfall in New Jersey without losing tropical characteristics was in 1903. Sandy continued to track WNW into Pennsylvania overnight while rapidly weakening, with the remnant low pressure stalling before drifting northeast, not exiting the region until early November.
Forecasting Hurricane Sandy:
The first model to show hurricane Sandy affecting the region was the CMC model, with the 10/20 0z run posted below (source: PSU e-Wall) – click on the image to enlarge.
Longer Range: Hurricane Sandy was a remarkable achievement for some of the forecast model guidance which has identified nearly 10 days in advance the potential for a historic East Coast hurricane. The first model to show such a solution was the Canadian (CMC) model with the 10/20 0z run*, 2 days before Sandy even became a tropical cyclone, showing a hurricane landfall over Long Island and New England.
At that time, it appeared to be a complete outlier as the CMC had a history of showing extreme, if not almost unrealistic solutions in its longer range, although it continued to show the extreme storm throughout the next few days, and was ultimately joined by the ECMWF model on its 10/21 12z run, and the GFS with the 10/21 6z run, at which point the three major models all indicated a highly unusual scenario of a hurricane tracking northwest into the region with heavy rain, extreme winds, and heavy snow. This was mentioned with the October 21 forecast discussion; the potential for a significant storm was noted in the forecast discussion, but with Sandy being 8-10 days away at that time, when forecast confidence is typically very low regarding specific details, this was simply noted as a potential. These signals continued on the 10/22 runs of the ECM and CMC, while the GFS was mostly further east with Sandy generally offshore.
The GFS model’s 10/24 6z run posted below (source: NCEP Model Analysis and Guidance) suggested that Sandy would remain well offshore, with a separate low pressure somewhat related to Sandy affecting the region. Click on the image to enlarge.
Medium Range: On the 23rd and 24th of October, the GFS and CMC models lost some of their initial support for hurricane Sandy to strike the region, while the UKMET sided with a further east track as well. At some point, both the GFS and CMC suggested that Sandy would track well offshore with a part of Sandy splitting and becoming a somewhat separate system affecting the region. The GFS model, however, was consistent with tracking Sandy well east of the region, either escaping well east of North America or tracking towards Atlantic Canada.
Despite the CMC, the first model to show direct impact from Sandy, losing its initial support, the ECM model remained consistent with Sandy making landfall in the region with extreme impacts, with a track close to the actual result, alternating between Long Island and New Jersey. Confidence was not certain at that time regarding a direct Mid Atlantic landfall, although confidence was gradually increasing, especially with the GFS and CMC individual ensemble members contradicting the operational runs and siding with the ECM; the October 23 forecast discussion noted these differences between the ensemble members and operational models, and while confidence was not high enough to forecast with certainty a direct hit on the region, confidence did increase on such a solution and the GFS was considered to be an eastern outlier.
The GFS’ model 10/27 0z run, showing Sandy making landfall over New Jersey, is posted below (source: NCEP Model Analysis and Guidance). Click on the image to enlarge.
Short Range: By October 25, four days before Sandy made landfall, the model guidance had settled on a consensus that Sandy would strike the region as a historic storm, while the ECMWF remained consistent with its earlier outlook of a direct landfall in the region. The GFS took longer to trend towards this solution, initially showing Sandy making landfall over Maine, not showing this taking place close to the NYC area until the 10/25 12z run. By that point, the sharp turn to the west and landfall in the Mid Atlantic was expected to happen, but the location of the landfall remained uncertain; aside from a handful of CMC runs taking Sandy into Maine, potential landfall locations generally ranged from Long Island or NYC on the GFS to the Delmarva Peninsula on the ECM, in which case the worst of Sandy would have stayed south of the NYC area. This range gradually narrowed down until the day before the storm, when
the consensus was settled for a southern New Jersey landfall.
On October 25, roughly 4 days away from the storm, it was just about certain Sandy would take the highly unusual step of retrograding west and making landfall over the Mid Atlantic region. The October 25 forecast discussion sided with this expectation, with heavy rain, coastal flooding and damaging wind gusts expected, although by that point the main areas of uncertainty were regarding the exact landfall location. There was high confidence regarding a general landfall range between the Delmarva Peninsula and Long Island, with this range gradually narrowed down until the day before the event. Confidence remained high on a landfall to take place over New Jersey during this time frame. The landfall location ended up in southern New Jersey, south of what the GFS model had indicated but north of the ECMWF, which took Sandy into the Delmarva Peninsula.
* The CMC and ECMWF models run twice a day, with an 0z run (1 AM) and a 12z run (1 PM). The GFS has four runs day, at 0z (11 PM the preceding day), 6z (5 AM), 12z (11 AM), and 18z (5 PM).
Impacts in the Northeast US:
Radar image of Sandy at 2:18 PM 10/29, approaching landfall
Cloud cover associated with hurricane Sandy spread into the region as early as October 27, while Sandy was still off the Florida coast. As Sandy continued to slowly approach the region, winds began to steadily increase on the 28th, with coastal flooding reported in parts of the Mid Atlantic regions over a day before landfall. The NYC area stayed mostly dry during this time period through early on the 29th other than scattered showers at times; meanwhile, heavy rain was already affecting southeastern parts of the region, especially in the Delmarva Peninsula, on the 28th, intensifying and spreading into Virginia and Washington DC by the morning of the 29th.
Radar image of Sandy at 7:18 PM 10/29, just before making landfall
By the afternoon hours on the 29th, as Sandy accelerated towards New Jersey, widespread heavy rain covered the Mid Atlantic region, with scattered rain squalls spreading further north into the NYC area and New England. An arctic air mass was in place over West Virginia, resulting in widespread heavy snow with near blizzard conditions developing in the western end of the storm. Winds steadily increased across the region, especially over New Jersey into the NYC area, reaching the 60-75 mph range over Long Island towards 6 PM.
The worst of the storm took place during the evening hours, however, as Sandy made landfall just southwest of Atlantic City around 8 PM. A strong low level jet moved through the area, with the strongest winds of the storm taking place during this time frame out of the southeast, coinciding with the high tide and a full moon to result in a significant storm surge in coastal areas of New Jersey into Long Island and southern New England. Widespread wind gusts of 60-80 mph were reported across the NYC area, with a gust of 90 mph reported in Islip, NY. As Sandy continued to slowly track inland into Pennsylvania, winds began to gradually weaken through the overnight hours, with windy conditions and scattered showers from Sandy’s remnants continuing for the next several days.