Thursday, November 15, 2012

Hurricane Sandy NWS Assessment Terminated

Well, that was fast.

The entire National Weather Service's "Service Assessment" for the "Hurricane Sandy Event" (as the assessment team named the study, hereinafter SA) has been terminated.

In an email at 9:46 this morning, Douglas Young of the National Weather Service wrote the SA team members:

I am writing to inform you that effective immediately we are terminating the spin-up of the National Weather Service Sandy Service Assessment Team.

We have been informed that a larger, multi-agency review of this event may take place... [emphasis mine]

Sandy will likely end up causing $100 billion or more in total (direct and indirect) damage and more than 125 deaths. Because of its importance, on November 2, I urged the NWS to charter a completely independent investigation.

The NWS, instead, decided to go ahead with its internal service assessment but for the first time put an outsider -- me -- in the co-chair position. I was told by David Caldwell, Director, Office of Climate, Water and Weather Services, that I would learn the NWS wanted an unbiased and thorough examination of Sandy.

Given the scores of deaths and the huge level of damage (according to media reports 100,000+ are still without power), even with excellent forecasts the Sandy Assessment may have been the most important the National Weather Service has ever conducted. Now it has been stopped. Why?

In the short period of time I worked on this project, a number of interesting questions surfaced:
  • Was there a decision not to call Sandy a "hurricane" regardless of its meteorological characteristics? If this decision was made, was it made Friday (October 26th) or Saturday morning? If so, who made the decision and why?
  • Was this decision the reason hurricane warnings, in spite of a large and dangerous hurricane moving toward the coast, were never issued?
  • Given that an obvious large and powerful hurricane was headed for the U.S. coast, why wasn't that decision reconsidered? For example, Barry Myers, the CEO of AccuWeather, urged (on the AccuWeather.com website) the immediate issuance of hurricane warnings about eight hours before landfall. Others also urged the lack of hurricane warnings to be reconsidered.
While the team's work officially didn't begin until Tuesday (November 13), I had already gathered hundreds of pages of information, literally dozens of suggestions, and some rather tantalizing leads that Sandy may or may not have been handled as it should have been from a scientific, administrative, or communications standpoint.

Now, in spite of all of the suffering and billions in damage, we may never know how well all the storm was handled.


A personal note: I was very impressed with the other members of the SA team. I'm sorry we will not be able to complete our assignment. 

22 comments:

  1. Mike, can you scan in those pages of info?

    ReplyDelete
  2. I've very puzzled by this blog post. We were told for a week that a 'hurricane' was headed up the coast. I've seen disputes as to whether Sandy was actually a hurricane or not when it made landfall, but it was known that 'Frankenstorm' was coming by everyone who didn't live under a rock. And in any case, while Sandy covered an exceptionally large area, it was not a particularly strong storm. As far as I'm aware, Sandy was a borderline Cat1 storm. The term 'hurricane' may have insurance implications, but for the general public at the time, there was no mystery. And the most serious damage in New York City was a function not of the 'hurricane' rating of the storm, but of the geometry of the New York Bight, which funneled the storm surge right into New York Harbor. If Sandy had made landfall in Boston, we'd be talking about a lot of rain, and power lines down. You make this sound like some nefarious conspiracy, when - even if it was - I can't see how it matters. New England takes a battering every year from Nor'easters, with much less than hurricane winds.

    ReplyDelete
  3. Could it (the investigation) have been stopped because if it is declared a hurricane then the insurance companies would have to pay out?

    I'm hearing from many people here in NJ... their insurance wont cover anything because it was once declared a hurricane and down graded to a tropical storm. Some even called it a cyclone. They can't categorize it except for calling it an "Act of God." And the insurance companies are using this as a way to not cover their clients. :-(

    ReplyDelete
  4. The Saffir-Simpson Hurricane Scale must be done away with.

    ReplyDelete
  5. @Rob: The NWS has wasted a week of my life (still have my regular job to do, was working till 1am on SA work multiple nights). I'm not going to scan documents.

    That said, if you would like to see some of the questions and comments for the SA take a look at my request to readers to ask questions for the SA team: http://meteorologicalmusings.blogspot.com/2012/11/hurricane-sandy-life-takes-ironic-turn.html

    ReplyDelete
  6. I understand, I was just hoping that what you found could be examined and utilized by the weather community at large...

    ReplyDelete
  7. Hi Mike,

    I'm sorry that your time has been wasted and that the American public won't get the benefit of the report that you and the team would have produced.

    I know in the grand scene of things that a week isn't that long to conduct and compile research, but would you consider a "Sirens" type work in the future on Sandy?

    ReplyDelete
  8. @PointSpecial and Rob, I'm still in shock. Not sure what, if anything, I will do about Sandy going forward.

    ReplyDelete
  9. Mike,

    From the infectious disease crisis / disaster point of view, this is our opinion, which agrees with yours:

    WHEN SIGNIFICANT LOSS OF LIFE OR SOCIO-ECONOMIC DISRUPTION OCCURS, THE UNITED STATES GOVERNMENT SHOULD FUNCTION UNDER MANDATE TO SERVE THE PEOPLE OF THE UNITED STATES WITH AN UNBIASED ASSESSMENT OF WARNING, WARNING-RESPONSE COUPLING, AND RESPONSE.

    The fact this does not occur in the public health world is a failure of governance in of itself. The fact that NWS appears to be attempting the same tact as well may be worthy of a bipartisan Congressional investigation.

    What usually happens when there is this large of an impact is avoidance of an open investigation by those in the elected leadership. A pattern observed over and over again. While NWS may have decided to proceed in good faith, it may be the leadership has intervened... the question is, for what purpose?


    James M. Wilson V, M.D.

    Chief of Station
    Ascel Bio National Infectious Disease Forecast Center

    Founder & Managing Partner
    Ascel Bio LLC
    jwilson@ascelbio.com

    ReplyDelete
  10. James - as has been pointed out in other threads, the SA will go on in one form or another. Rumor is that the move from DOC is a possible delay, but not cancellation.

    ReplyDelete
  11. "the Sandy Assessment may have been the most important the National Weather Service has ever conducted. Now it has been stopped. Why?"

    Maybe they read "Sirens" and got cold feet?

    After all, the entities that would be scrutinized this time around have even LESS excuse for their lapses/failures than SGF, SPC, local EM, etc. did in Joplin.

    With Sandy, all involved had DAYS of advance warning and multiple reliable forecast models agreeing (after some initial divergence) on where the storm would track. So how would they justify, for instance, not issuing hurricane warnings for a storm that looked, walked, and quacked like a hurricane?

    Elaine

    ReplyDelete
  12. @Elaine: I think you are very much in the ballpark.

    FYI: The acting director of the NWS, the chief scientist at NOAA, and the head of NCEP all had copies of "Sirens" well before Sandy came along.

    ReplyDelete
  13. I still wonder if Hurricane Warnings would have made a difference. I say this in part from the fallout of Irene and the over forecasting of this storms intensity on the coast.

    If the high wind warning were questioned why not issue or create ad hoc a High Wind Warning for Hurricane Force winds? The NWS is now using enhanced tornado warnings (Tornado Emergency). Why not issue an enhanced High Wind Warning?

    Also too another issue was the fact that the worst coastal flooding occurred at the time of high tide and the surge. While surge was correctly forecast the storm flood tide was higher 13+ feet over the NY Bight. Advisories should have forecast maximum flood waters or potential maximum flood heights not the surge.

    Someone hears 6-11 feet storm surge public perception is oh 6-11 feet of water will come onshore. Also why not include communicative benchmarks in the coastal flood warning like the NWS issues with river flood warnings. e.g a river is forecast to go into MAJOR flood then the flood advisory mentions at major flood the water plant at such and such location is undated or this road or that is cutoff.

    Do the same kind of BM with the coastal flood warning: at 6.5 feet water enters the PATH Tubes at 7 feet they flood and water enters the trans-Hudson Tunnels. At 7.5 feet water spills over the Battery, etc.

    ReplyDelete
    Replies
    1. I think it would have made a difference... Bloomberg waited an extra 24 hours to order an evacuation... And did so only for zone A. If you listen to his press conference from Saturday (before the storm), it was clear he was getting intelligence that it would not be as bad as a hurricane or tropical storm. Plus, by definition, hurricane warnings automatically trigger certain evacuations and emergency steps that were not taken. In this case, time wasted = lives lost in my opinion.

      Delete
    2. @Andrew: Please see my post in Mike's original blog on the assessment team's ask. To review: NWS' Meteorological Development Laboratory and the WFO's indeed have the capability to produce both storm surge INUNDATION forecasts, and potential impact wording that provides "real world" language for what a high to extreme level of impact can do (i.e. "water up to the level of a two story home"; "vehicles washed away", "Hundreds of structures inundated or washed away", etc). Both of these functions were used effectively during Isaac. My hunch is that the lack of a hurricane warning was reason for the local offices and MDL not to issue these information streams - which may have helped clarify decisions.
      http://weather.gov/tcig (click on an office for potential impact wording...suggest Brownsville or New Orleans for best examples)
      http://www.nws.noaa.gov/mdl/phish (Probability of Hurricane Inundation from Surge Height).

      Delete
  14. You and the other members should continue to work without the NWS. I am sure we can all work on something amazing regardless...

    ReplyDelete
  15. "Could it (the investigation) have been stopped because if it is declared a hurricane then the insurance companies would have to pay out?"

    I don't think so. From what I understand (I live in the Midwest so I don't have to worry about this, but a lot of other people do), the common practice nowadays is for insurance policies to have a much higher deductible (usually a percentage of the home's total value rather than a flat dollar amount) when the damage is from a "named storm", which includes both hurricanes and tropical storms. As long as Sandy was a "named storm" that clause would apply.

    Elaine

    ReplyDelete
  16. Readers: I'm impressed at the high quality of discussion here. Thanks.

    ReplyDelete
  17. @Barry TY for the post links.
    @Jason Good point
    @Bookworm I have a few friends in CT and on LI and their deductibles go up by 20% based if the warning is a tropical warning NOT a HIGH Wind warning

    ReplyDelete
  18. still interested in following up on the sirens reference, if anybody can clue me in.

    ReplyDelete
  19. My book, "When the Sirens Were Silent."

    ReplyDelete