Sandy Claus was coming to town

The precautions against superstorm Sandy worked because the US government FEMA listened to a European computer forecast, not its own.

We need something seasonable to cheer us up, not contemplating Republicans in Congress, narcotics, guns, gun apologists, and – far the worst – the faces of small children. So let’s take a short trip down memory lane to your friendly ex-hurricane:

Tropical storm Sandy at midnight on 28 October

Source: NASA via The Telegraph, approximate scale bar added by me

Where is the silver lining? Two actually. The first is the fact that Sandy was about the smallest storm still large enough to shift public and élite opinion on the reality, now, of climate breakdown. If you pay attention to the plague of frogs, you just may escape the cull of the firstborn. The second is much less widely known outside the weather forecasting trade, and a reminder of just how important it is.

The NOAA´s own computer model, GFS, predicted that Sandy would stay out to sea. [Update : link to article by Andrew Freedman at Climate Central added.] Very fortunately for the USA, it wasn´t the only game in town. An intergovernmental European weather forecasting organisation, the European Centre for Medium-Range Weather Forecasts with the handy acronym ECMWF (Soviet Russia could not have done a worse branding job), has its own model of the Atlantic weather patterns. It needs to, because these determine European weather a few weeks later. Their model correctly predicted that Sandy would turn inland, and be violent. A contemporary insider blog post by Fred Campagna on October 23, six days before Sandy made landfall, shows that the different perspectives were well known in the trade at the time:

If there is one model to blame for all the Sandy hype, it’s the ECMWF. This model has been consistently bringing a major storm, 950mb or lower, to the Northeast. … The ECMWF has been consistent, but it is still farther west than most other models. The critical juncture for determining whether it is correct will likely be Friday or Saturday when the storm is in or leaving the Bahamas. While I think the GFS is too far east, I think the ECMWF may be too far west, and I’m nearly certain it is too intense with the storm.

FEMA went with the European forecast in its thorough planning. 125 lives were lost in the USA, with about $63 bn of property damage. The disproportion between the two figures indicates that lifesaving precautions, including evacuation, worked well. Hundreds, possibly thousands, of lives must have been saved. There will also have been some reduction in property damage from the warning; a modest 5% would still mean $3bn.

European taxpayers (including me) and international officials and scientists at the EC.. (let´s just call it the Reading Centre) are therefore owed a very pretty Christmas card of thanks from Bloomberg, Christie, Fugate and Barack Obama.

We do not expect a billion-dollar cheque. Europeans need the Atlantic forecasts for our own purposes, and you are welcome to the spinoff benefits. It´s called comity; decent people help each other out. We are also rather chuffed that our model is better than yours. Come and have a look.

You can do one thing, for all of us. The Reading model uses data from American weather satellites. (Thank you, US taxpayers.) One of these, a polar satellite called Suomi NPP, is running out of time for replacement. Reading have run a simulation  of the model without the polar satellite data – and it gave the same wrongly optimistic prediction as the American one. They are too scrupulous to say that this would always happen. Still, disaster planning is all about the precautionary principle. There´s far too much at stake to do without the satellite. It shouldn´t wait to 2017 as currently planned.


Meteorologists can´t predict hurricanes years in advance, but they have already given them names. Here´s the list of the potential  North Atlantic tropical cyclones to 2017:

Andrea, Barry, Chantal, Dorian, Erin, Fernand, Gabrielle, Humberto, Ingrid, Jerry, Karen, Lorenzo, Melissa, Nestor, Olga, Pablo, Rebekah, Sebastien, Tanya, Van, Wendy
Arthur, Bertha, Cristobal, Dolly, Edouard, Fay, Gonzalo, Hanna, Isaias, Josephine, Kyle, Laura, Marco, Nana, Omar, Paulette. Rene, Sally, Teddy, Vicky, Wilfred
Ana, Bill, Claudette, Danny, Erika, Fred, Grace, Henri, Ida, Joaquin, Kate, Larry, Mindy, Nicholas, Odette, Peter, Rose, Sam, Teresa, Victor, Wanda
Alex, Bonnie, Colin, Danielle, Earl, Fiona, Gaston, Hermine, Ian, Julia, Karl, Lisa, Matthew, Nicole, Otto, Paula, Richard, Shary, Tobias, Virginie, Walter
Arlene, Bret, Cindy, Don, Emily, Franklin, Gert, Harvey, Irma, Jose, Katia, Lee, Maria, Nate, Ophelia, Philippe, Rina, Sean, Tammy, Vince.


I´m glad that my name isn´t on the list. If yours is, cross your fingers that your homonym won´t be the first Big One coming to town. Which coastal city will it strike? Miami? Savannah? Norfolk? Boston? Or New York again?

A cheery and grounded Christmas (or other midwinter festival of your choice) to you all.

Author: James Wimberley

James Wimberley (b. 1946, an Englishman raised in the Channel Islands. three adult children) is a former career international bureaucrat with the Council of Europe in Strasbourg. His main achievements there were the Lisbon Convention on recognition of qualifications and the Kosovo law on school education. He retired in 2006 to a little white house in Andalucia, His first wife Patricia Morris died in 2009 after a long illness. He remarried in 2011. to the former Brazilian TV actress Lu Mendonça. The cat overlords are now three. I suppose I've been invited to join real scholars on the list because my skills, acquired in a decade of technical assistance work in eastern Europe, include being able to ask faux-naïf questions like the exotic Persians and Chinese of eighteenth-century philosophical fiction. So I'm quite comfortable in the role of country-cousin blogger with a European perspective. The other specialised skill I learnt was making toasts with a moral in the course of drunken Caucasian banquets. I'm open to expenses-paid offers to retell Noah the great Armenian and Columbus, the orange, and university reform in Georgia. James Wimberley's occasional publications on the web

5 thoughts on “Sandy Claus was coming to town”

  1. NOAA and the NHC typically use an ensemble of models to generate a “consensus” model, then the forecaster adjusts that based on experience and judgement. UKMet follows a similar process. There are currently about 10 models that generate fairly good results, but each one tends to do better for certain types of hurricanes, certain ocean areas, certain weather patterns, etc. And the models tend to have good and bad months or years for reasons which are not understood. So personally I would say it is a bit misleading to say “The NOAA´s own computer model, GFS, predicted that Sandy would stay out to sea. Very fortunately for the USA, it wasn´t the only game in town. An intergovernmental European weather forecasting organisation, the European Centre for Medium-Range Weather Forecasts with the handy acronym ECMWF”.

    I will note that the beloved former Senator Santorum was pushing very hard in the 2004 timeframe to terminate the National Weather Service and most of NOAA and sell their data to Accuweather, a private forecasting firm. Very fortunately that didn’t happen and Santorum is no longer polluting the Senate.


    1. The source for my statement about GFS is the penultimate link; I’ve added it to the sentence you comment on. The Climate Central article by Andrew Freedman says: “The NWS’ own GFS computer model failed to accurately predict the path of Hurricane Sandy five days in advance, a marked contrast to the European model’s performance.”

      Your clarification is helpful, but looks like a nuance more than a correction. Fred Campagna, the blogger I cited, wrote that the European model was an outlier. Were the White House and FEMA really only getting a “consensus forecast” with a single probability of landfall, or reports of the clear divergence between the models which my citation indicates? On something this important, I’d expect the forecasters to have to explain the methodology and the variance round the central estimate, as with climate change. A 25% chance of major impact would be enough to get action from most of us.

  2. Not sure how to add a link (are there instructions anywhere?) but here’s a more detailed but still very readable explanation of how the ECWMF model got a one or two-day lead on the others: The key factor, as James mentioned, was not that the ECWMF model is intrinsically better than the others, but that it incorporates more data, particularly from polar satellites. And yes, we have a current and large problem with funding for those.

Comments are closed.