News•May 15, 2013

Panel Finds Flaws with NWS Guidance on Sandy's Surge

The National Oceanic and Atmospheric Administration (NOAA) released its final review of the agency’s performance during Hurricane Sandy, finding that while the agency provided accurate forecasts of the storm’s path and strength well ahead of time, there were many shortcomings, including confusing and poorly timed guidance on the expected storm surge.

Pieces of the boardwalk float in sections through the flooded streets of Atlantic City.
Click image to enlarge. Credit: billmckibben/twitter

It was that record storm surge that caused most of the damage along the New Jersey coast and into New York City, where it inundated every subway tunnel between Manhattan, Queens, and Brooklyn, and helped cut power to lower Manhattan for days. The surge was also the dominant cause of storm deaths, with 41 victims, the report said.

In total, the storm was responsible for 147 direct deaths, the destruction of at least 650,000 homes, and it caused at least $50 billion in damage and left about 8.5 million customers without power.

The assessment team, which included meteorologists, emergency management officials, and social scientists from NOAA, FEMA, and the Centers for Disease Control and Prevention (CDC), focused especially closely on the National Weather Service’s (NWS) forecasts and warnings, including the policies underlying the warnings and watches that were issued; the efficacy of NWS websites as portals for storm information; and the production and issuance of storm surge-related forecasts.

The assessment praised the NWS for its advanced notice of the storm’s threat, and for emphasizing its potential to cause significant damage. The agency first highlighted the potential for the storm to bring major coastal impacts with nine days’ lead-time. The forecasts the agency issued, including the surge forecasts, were remarkably accurate, the report found, but it found fault with the ways the forecasts were communicated.

In particular, the assessment found that the storm-surge guidance provided by the agency’s various offices were often confusing and lacked any graphical presentation. The storm surge caused coastal flooding to exceed 8 feet above ground level in some areas, including in portions of New York City, and water rise was noted from Florida to Maine, the report said.

Ocean City, New Jersey.
Click image to enlarge. Credit: kenshane/twitter.

The report identified the urgent need for high-resolution storm-surge forecasting and communication as the “highest priority need” to emerge from the Sandy review process. “In particular, there is a crucial need for storm-surge graphical inundation guidance,” the report said.

In addition, the report said that the issuance of non-tropical watches and warnings did confuse some emergency managers, including the New York City Office of Emergency Management. That finding lends credence to the possibility that the lack of a hurricane warning, along with confusion over Sandy’s storm-surge potential, played a role in the actions of New York Mayor Michael Bloomberg, who initially downplayed the storm’s potential impacts in a press conference two days before the storm made landfall, only to reverse course and order the evacuation of low-lying areas the next day.

“(The) NYC Office of Emergency Management (OEM) was confused about the storm-surge forecast, what and where the impacts would be, and how high the water would rise,” the report said.

After a press conference on Oct. 27, at which Bloomberg had downplayed the storm’s threat, in part because it was not predicted to be a hurricane at the time of landfall, staff at the National Hurricane Center (NHC) called the OEM directly. According to the report, the city’s emergency managers considered the call to be “an eye-opener” about the potential storm-surge impacts in the city.

“This apparent communication breakdown may have delayed critical decision making about evacuations and led to potentially confusing messaging to the public,” the report said

“It is not expected to be a tropical storm or hurricane-type surge,” Bloomberg said on Oct. 27. “With this storm we're likely to see a slow pileup of water rather than a sudden surge which is what you would expect from a hurricane and which we saw with Irene 14 months ago, so it will be less dangerous but make no mistake about it there will be a lot of water and low-lying areas will experience flooding.”

In response to the NOAA report, the OEM said the storm surge forecast for New York changed between Oct. 27 and 28, which is why the evacuation was ordered on the 28th.

“NOAA's characterization of OEM and other Emergency Managers as 'confused' in its review of the surge data provided by NWS fails to address a key fact: The NWS surge forecast provided to OEM changed significantly between October 27 and 28 — from 4-8 feet to 6-11 feet. This was the key driver of the city's decision to order an evacuation on October 28,” an OEM spokesperson told Climate Central.

The report found that even NWS personnel were confused by the ways in which NOAA measures storm surge, such as the height of water above the often-cited Mean Lower Low Water level.

Perhaps most importantly, the report found that the NWS and media were unable to translate the storm-surge forecasts from mere numbers to effective graphics and impact information.

“The public and NOAA/NWS partners did not clearly understand what storm surge was or how dangerous it could be,” the report said. “Most of the media representatives the Service Assessment Team interviewed believe NWS is not giving them any means to communicate predicted surge impacts to the public.”

One anonymous media representative quoted in the report said, “There were forecasts, but there were no details about what the predicted surges would cause . . . There was no wording in the warning like, 'Most of lower Manhattan will be underwater at a certain point' or, 'Staten Island will flood up to Beach Street.' ”

NWS Director Louis Uccellini said he intends to implement the report's 23 major recommendations for service improvements. “I was very impressed with the thoroughness of the job that they did,” Uccellini said in an interview regarding the assessment team's work. “We’re going to appraise the report, work with it, and implement it,” he said.

A GOES-13 satellite image taken on Oct. 29, 2012 shows Hurricane Sandy centered off of Maryland and Virginia.
Credit: NOAA

The report also cited a March 2013 poll that found that 79 percent of coastal residents surveyed said the storm-surge impacts in their area was “more than they expected.”

In response to Hurricane Sandy as well as other recent storms, such as Hurricane Isaac in 2012, the National Hurricane Center (NHC) is currently developing storm-surge warnings and plans to roll out new inundation maps on a trial basis as soon as the upcoming 2013 hurricane season, which kicks off on June 1. The storm-surge warnings, however, will take at least another year to test and roll out.

The report found the NWS is impeded by a staffing shortfall in its storm-surge unit. Currently, the agency has just one full-time federal storm-surge forecaster at the NHC in Miami, along with two private sector storm-surge model developers.

The assessment found that the NWS made the decision not to issue hurricane warnings north of North Carolina due to the strong preference from the emergency management community, and at least one governor, that the warning type not change even if the storm were to transition from a tropical storm or hurricane into a post-tropical cyclone, as computer models predicted it would do.

Under policies in place at the time — which have since been changed — if a hurricane were to transition into a post-tropical cyclone prior to landfall, then local Weather Service forecast offices would be given responsibility for issuing non-tropical watches and warnings, rather than continuing to have hurricane warnings in place and coordinated from the National Hurricane Center in Miami.

In Sandy’s wake, NOAA has revised this policy, and has given the NHC the option to continue to keep hurricane watches and warnings in place even after a storm has transitioned into a post-tropical stage.

This new policy will enable NOAA to be consistent in the way it names tropical storms and hurricanes and issues warnings for them, and will prevent situations like Sandy in which forecasters are presented with an array of less-than-ideal options, each with the potential to confuse the public about how serious a storm really is.

Critics of the NWS’ decision not to issue hurricane warnings have contended that perhaps more people would have evacuated ahead of the storm, had those warnings been issued.

However, the assessment report rebuts that line of criticism by citing social science research showing that the lack of hurricane warnings did not cause people to react with reduced concern. Based on post-storm surveys, the report said that many people actually thought they were under hurricane watches or warnings when the storm struck, and that their level of concern “did not differ substantially” from respondents who thought a different type of warning was in effect.

“Many vulnerable residents ignore evacuation orders, even when they are accompanied by hurricane watches/warnings,” the report said.

While singling out the NWS New York and New Jersey offices for praise, the assessment faulted the NWS for the difficulty users had in finding critical information easily on the agency’s myriad different websites and social media platforms, noting that NOAA websites received close to 1.3 billion hits during the storm.

The team found that the heavily trafficked Hurricane Center homepage, for example, did not contain the flood and high wind warnings and storm bulletins from local forecast offices, as well as storm-surge information, issued by each office. The lack of a one-stop-shop for online information about the storm was problematic, and the New York OEM told the assessment team that they relied on a non-NOAA website for storm-surge guidance, which they found easier to use.

The report found that the NWS is ill-positioned to take social science insight about how people process and respond to storm-related information into account when wording its warnings and designing its graphics and websites, which can lead to ineffective products.

The NWS, the report said, has only one employee with doctoral-level training in a social science field, and the underrepresentation of social science expertise within the agency has hindered the effectiveness of the agency’s risk communications, especially during major events such as Sandy.