I've twice read this National Hurricane Center introspective on Tropical Storm Erika by James Franklin, NHC's Forecast Branch Director. The write-up is excellent, and a must read for anyone in the business of communicating tropical cyclone threats (not just broadcast meteorologists, but reporters, news directors, producers, etc.). As I was reading it, it seemed like Mr. Franklin was speaking directly to me, as he addressed many of the questions and critiques I posed on air and in social media during and after the storm. He also mentions my use of wind speed probabilities in communicating the risk of Erika affecting South Florida, which I'm proud of.
I applaud the frank exposé of how the NHC forecast accuracy was off. While I concur that the models are increasingly difficult to beat, I disagree with the self-admitted nearly-blind reliance on them. I mean, those 12 and 24 hour NHC forecast positions (resulting from over-dependence on the models) that depicted Erika's "hard-right turn" into Puerto Rico and Santo Domingo were completely unrealistic. And this led to even greater 3, 4 and 5 day errors in track and intensity. I'm troubled by James Franklin's concern that "our partners and the public might rightly have questioned our rationale to go firmly against many model forecasts of a stronger system farther north." In my opinion, meteorologists specializing in forecasts -- including severe weather forecasts -- are there to make tough calls, not safe calls. Each tough call needs to be accompanied by a robust set of supporting facts and trends, as well as a depiction of confidence level (both in simple terms and in more complex probabilistic data), plus alternate scenarios should the confidence in the main idea be low. This information should be included (as suggested in Franklin's post) in more than one place, not just the technical Tropical Cyclone Discussion.
I share James Franklin's disappointment that more of us in charge of guiding the audience through tropical threats are not picking up on (and broadcasting) all the clues that NHC is providing in plain sight to convey the true risk level. Franklin explicitly urges NHC's media partners to "take some responsibility for making use of the information that is already available." To this I add: communicating the possibility of a storm or hurricane affecting South Florida isn't just about a cone and where Miami sits in relation to such cone.
The cone has become the ON-OFF switch for media hysteria and coverage (something which I work very hard at to tone down as much as possible here at NBC 6 South Florida). How about we implement a new interactive cone graphic which *always* includes wind speed probabilities? Imagine if when the forecast included a Category 1 Hurricane Erika hitting Miami, the actual 4 or 5% probability of hurricane-force winds was displayed right over the city? That would lower the volume of the hype, don't you think?
