make WWAY your homepage  Become a fan on facebook  Follow us on twitter  Receive RSS Newsfeeds  MEMBERS: Register | Login

BUCKLEY'S BLOG: Why a perfect forecast just isn't enough

The TV meteorologist has long been the butt of many jokes. "What's the only job you can have and be wrong everyday? A meteorologist of course!" (Har, har, har.) Why is this the case? Well, as meteorologists we're trying to predict the future.

Great forecasts in history. (AP Photo)

That can a tough business sometimes. Ask the economist, putting out a seasonal outlook for gas prices. Or maybe the political analyst, confidently explaining who will win in November and why. Or the stock broker, making sure your 401k is never losing a penny. Inherently, predictions are a precarious, and often perilous pursuit. But things are changing in meteorology.


More than ever before, your daily forecast is an accurate portrayal of the weather you can expect to impact you that day. A 5 day forecast today boasts the accuracy of a 2 day forecast 20 years ago, and it's getting better every day. Does this mean meteorologists of today are smarter than those of the decades before? Probably not. Instead, the advent of computing power has been advancing the science of meteorology at hyper-speed with no signs of slowing down. These "computer models" you hear about from time to time during a big weather event really are something, and they help us get it right more than ever before. In fact, I think we're now on a path where getting the forecast right will be secondary to how well we communicate it.

As broadcast meteorologists, we have tricky jobs. To excel, we must be able to consistently do two things well. One, create an accurate forecast. Two, communicate that forecast and its consequences in a clear, easy to understand, impactful way to a wide ranging audience. You have to be great at both! Think about it. What good is a great forecast if people don't understand the impact? That question has been rolling through my mind over and over after seeing the devastation of Hurricane Sandy on the Jersey Shore and in New York over the past week. 

30 miles accuracy from 5 days out. (Click for larger image. )

The forecast for Hurricane Sandy was nearly flawless. The National Hurricane Center nailed the landfall location within 30 miles 5 days out -- 5 days!The high performing European (ECMWF) computer model had the idea of an historically unprecedented track a week ahead of time, the NHC jumped on board, the storm surge was predicted, and the forecasters wrote themselves a success story. But as I've said, that's only one part of the equation. While the forecast was as good as we could have hoped for, I'm afraid the threats weren't as clear to the public as they were to us meteorologists. 

As I'm writing this blog, over 80 people have lost their lives in the US due to Sandy. That number would put Sandy as the 13th deadliest US tropical cyclone. We have to do better. So what went wrong? I certainly don't have all the answers, but here are some things I've noticed in the past few days that can start the discussion. 

Devastation in Seaside Heights, NJ after Sandy. (Image from NOAA)


1. Orders to evacuate coastal towns often went ignored: The forecast called for unprecedented storm impacts, particularly with storm surge. With ample time, officials such as Mayor Bloomberg and Gov. Christie made the right call to issue evacuations. For a number of reasons, the public simply didn't get out of the way of the storm. In the end, the evacuation areas were the ones hit the hardest. We need these orders to be taken seriously. Psychiatrist Dr. Keith Ablow offers some thoughts in this article, where he states we may be losing our grasp on reality. 

"In such times, a governor's real and heartfelt and sage warning to evacuate a coastal area may not overcome the background noise of reality shows." - Dr. Keith Ablow, Fox News.

2. The Irene problem: Last August, another huge storm was headed for New York and prompted evacuation orders. The problem was is that Irene wasn't as bad as the pre-storm hype / forecast led the public to believe. While the storm hit the interior Northeast hard with flooding, it did not pack a punch with surge at the coast and did not lead to any devastating damage in the evacuated coastal areas.

One resident of Hoboken, who is dealing with massive flooding now summed it up this way, "I feel like the storm was way more intense than how they were describing it. We over-prepare last time [for Irene]. This time, they underprepared."

Did last year's storm create a "cry wolf" scenario that was too tough to overcome, even with a great forecast? For some it appears so. 

3. It's difficult to prepare people for a storm they've never experienced. Quite simply, it's hard to explain what will happen to somebody if they have nothing to compare it too. Here in NC, we would easily be able to say, "This storm will be like a Fran", and people will instantly have mental images of what impacts to expect. In the Northeast, there hasn't been a coastal storm with these impacts in generations, if not ever. I wonder if people just couldn't imagine storms could do this type of damage.  

4. Hurricane Warnings vs. "regular" warnings? This is a big one. Before Sandy hit, there was much debate over whether it would be a "true" tropical system, or if it would achieve a hybrid "extra-tropical" status. This has to do with the science of whether the storm is warm-core or cold-core and how it gets it's energy to survive. In the end, the NHC decided that Sandy would transition into a post-tropical system before making landfall, and that it would be inappropriate to issue Hurricane Warnings for the coast even though the impacts would be the same. Instead, the local NWS offices issued High Wind Warnings / etc for the coasts.

To me, I feel like this was a real stain on an otherwise great effort by the NHC with this storm. The public is often inundated with products from the NWS, whether it be coastal flood advisories, wind advisories, etc. But, in New Jersey residents could certainly count on one hand the number of times they've ever been under a Hurricane Warning. This was a true missed opportunity to communicate the severity of the threat, and potentially get people to get out of the way. I think it was also an example of the science getting in the way of the message. 

5. Can we do better than the Saffir Simpson scale? How many times have you heard, or said, "It's only a Category 1." So far this year that has been nails against a chalkboard for me twice. The first time with Hurricane Isaac as it neared New Orleans. Folks in that hurricane zone didn't take the storm seriously because it was a 1 and they've been through many 1's before; what's the big deal? Well, Isaac was a larger than normal storm that would stall out on them and dump copious amounts of rain. As we know with Sandy (a storm that was "just a 1"), the impacts were enormous not from wind but with storm surge. Let's be honest, if Sandy actually was going to hit here, would our area in Southeastern North Carolina have taken it seriously as a Cat 1? (I'm thinking it would have been a tough sell)

Hurricane Sandy off the NC coast. (WWAY Photo)

With the current scale all we are rating with hurricanes is wind. Just one part of a storm! I would love to see a new scale created that is more impact driven. Instead of just looking for a high wind gust in the storm center, actually forecast the impact of the high winds, storm surge, inland flooding, size and scope, and rate the system accordingly. This would be a radically different scale, but wouldn't it more accurately communicate a threat to the public? After all, no two hurricanes are created equal. 

There are a lot of successes to be celebrated with Sandy, but there are also many lessons too. Probably more than any other science, meteorology must constantly communicate scientific information to the public. On a daily basis, it can be getting people to understand what the dewpoint is, but in severe cases it can be life threatening information. On the whole, we have a lot of work to do -- after all, "Scientists are lousy communicators" as stated in this Newsweek article. As forecasting skill continues to improve, I hope the community as a whole starts to focus more efforts how we reach the public so that our predictions aren't for naught. 

PS: I'd love your input! Living in a hurricane zone, we often need to communicate a storm's threat. How can we do this better? What information do you need, and how do you like it presented to you so that it makes sense. I want your feedback. If you have comments you can leave them here on this blog, shoot me an e-mail at, find me on Facebook at, or find me on Twitter @TimBuckleyWX.

Thanks for reading!

- TB

By: Tim Buckley

Excellent Article

Tim, excellent article that succinctly highlights the challenges that meteorologists face. I think that although NHC did a superb job with their forecast, they did a huge disservice by dropping the hurricane warnings. The public is largely ignorant. Explaining warm core vs. cold core is so far over the average person's head, that it is a waste of time and not helpful to the public. This is not Meteorology 101, people need to know the IMPACT of the storm, not the process that is generating it. I also agree that it is time for the Safir Simpson scale to go. Joe Bastardi has been working with a scoring system that utilizes pressure and wind speed to predict the intensity of hurricanes. That would be a good start. A storm impact scale similar to the Regional Snowfall Index developed by Kocin and Uccellini would be a great way to inform the public of the danger and expected impact of landfalling hurricanes beyond maximum sustained wind speed. I know that several papers and talks will focus on the meteorology behind Sandy and her impact from the Caribbean to the Northeast. I hope that at least some of those presentations can focus on a better way to present the potential risks to lives and property and that a new approach can be implemented by next hurricane season.

Communication equally important

Completely agree Scott. Thanks for the kind words. We have a long way to go in terms of making these complicated threats clear and plain as day to the public. That's not to say we don't have a lot to be proud of now, I just think we can make it so much better in the future.