There was no emergency in Florida, of course, but this was the headline on media outlets, both local and national. I asked Governor Rick Scott if “state of emergency” wasn’t inflammatory in a live interview on The Weather Channel last Friday. He gave a boilerplate answer, “we want everyone to be prepared”. Of course we do, but it’s the wrong answer. In fairness, every governor in recent history has made the same unthoughtful declaration because that’s the way the system is set up. A number of important preparedness processes kick in under a state of emergency, which are good and necessary.
So the state wants to be ready in case a storm comes its way. Great. But how about changing the official declaration to a “state of readiness”, or something equally meaningful, instead of using words that are obviously inappropriate to the situation. Words count. And inflammatory words must mean that something extreme is happening, or they won’t have any meaning when a real disaster comes along.
The hurricane-hype problem is related, but much more difficult. Watch the national news these days, and every storm is somewhere between hyper-dangerous and the end of the world. There is no excuse for it, except they obviously don’t ask an experienced meteorologist, or don’t want to know. If you believe the news, every event is “only going to get worse”.
The problem during Erika was exacerbated by the fractured system the National Hurricane Center uses to communicate their forecast. The official forecast is spread over five or more text bulletins and three or more graphics. Only the most experienced user knows how to aggregate an accurate understanding of the NHC’s thinking by assimilating the multiple links and formats, and even then the bottom line is often buried in the verbiage and/or a technical understanding of the products. Most media people, social and otherwise, are not experienced users or are overwhelmed by the mechanics of deciphering and communicating the message on deadline. The result is: the public suffers.
The fundamental structural flaw in the current system is that the hard-coded format controls the message. Read an advisory from Hurricane Donna 55 years ago and you understand the threat. Read an advisory today and you’re buried under the trees when the forest is what counts.
Scientifically speaking, the NHC had no business making 96- or 120-hour forecasts for Tropical Storm Erika last week. The state of meteorological science isn’t sufficiently advanced to make a deterministic extended forecast given the conflicting signals the models and common sense were sending. As the NHC hurricane specialists accurately said in their technical discussions, the forecast beyond 72 hours was low confidence, which was an understatement. But, the format required numbers – specific, confusing, hypeable numbers and locations for the center of the storm.
In fact, of course, they weren’t making forecasts as much as they were establishing reasonable worst-case upper limits on the winds that might accompany the storm four or five days in the future. Not only is there nothing wrong with conveying reasonable worst-case values when you have very little idea what’s going to happen, it’s the only responsible thing to do. The problem is the NHC conveys these wild-ass guesses the same way it conveys critical forecasts in which they have high confidence. The format constrains the communications, and the big numbers become the story.
As long as the format is hard-boiled so the forecaster has no ability to adjust the communications style based on the message he/she wants to convey – depending on the forecast confidence the science allows – the correct message will be hard to divine and the hype-inclined media will continue to distort the threat. Experienced meteorologists knew that Erika was heading into chaotic atmospheric and topographic territory, and that the variables were higher than normal. But, that wasn’t the topline message from the NHC.
In the end, of course, the NHC forecasts had the basic picture right. They said a tropical system was coming toward Florida, and it did. Hidden in the verbiage they said they didn’t know how strong it would be, and they didn’t. But because we expect every forecast to fit in the same mold and be evaluated the same way, even when the science doesn’t support it, Erika goes down as a forecast failure.
So there are two mountains to climb. The NHC needs to figure out a way to convey their confidence in the forecast and/or their ability to even make a forecast. Burying the explanation in the technical discussion is not sufficient in a 140-character world. (I know it’s not trivial and I actually have another idea, but I’ll save it for another blog.) And the media needs to quit sensationalizing weather events so that nobody believes anything. It wasn’t that long ago that real journalists would clarify a statement like “a state of emergency is declared in Florida”. Now it leads the national news.
No comments:
Post a Comment