Forecasting the weather is a unique job, especially in severe thunderstorm season. When severe weather is possible, we’re put in the position of giving people bad news leading up to the threat…and then having to sometimes apologize for delivering GOOD news when the threat goes down! Either we get the forecast right and our neighbors’ lives and property are in danger, or we get the forecast wrong and we end up with egg on our faces.
Yesterday’s severe weather threat obviously represents the latter scenario. Now, there were SOME severe weather reports — 15 wind reports and one hail report:
Most of the wind damage came from one storm that developed ahead of the “main line” and ripped up though Rutherford and Wilson counties just after sunset.
But I’m not here to say, “See? There was severe weather! We didn’t do so bad…” I’m here to take a look back at what we said leading up to Thursday, and evaluate what we can learn from yesterday’s thunderstorm-fizzle. I’m going to include some of what I wrote in blog posts leading up to the event — just to be clear, there’s essentially no difference between what I write on the blog and what I say on the air. I phrase things a bit differently (I can be nerdier here) but I emphatically do NOT believe in giving one forecast on-air and a different one online.
We obviously had a big severe weather event Monday evening — my first real look at Thursday’s severe weather threat came Tuesday morning, when I wrote this:
I’m FINE with that. It matched up with the data available at the time — just that short blurb gave people a heads-up, communicated the most-likely time frame, and communicated that there was still significant uncertainty.
But on Wednesday…oh, Wednesday. I woke up to see that the Storm Prediction Center had placed us in a Moderate Risk of severe weather. Here’s what I wrote in the blog that morning (the whole post is here):
All of that is factually true. But while the SPC generally does an excellent job forecasting severe weather events across the country, once in a while they do undersell or oversell an event. I was trading messages with one of the @NashSevereWX guys on Monday, expressing my frustration that SPC wasn’t aggressive enough with their phrasing for Monday evening’s threat. But on Wednesday, I didn’t spend enough time evaluating whether SPC was being too aggressive for Thursday’s threat. I included the SPC’s own model data (the SREF) in Wednesday’s post, which showed a 70% chance of thunderstorms with supercell characteristics. But a 70% chance, combined with the fact that the specific severe-weather indices were far from off-the-charts, should really point to an “Enhanced Risk” evaluation, rather than a “Moderate Risk.”
I didn’t mention ANY of that, nor did I indicate any level of uncertainty on the map above. Television meteorologists have an obligation — that is, we are required — to pass along watches from the SPC and warnings from the National Weather Service when they are issued. (There are a couple of stations elsewhere in the country that go off the rails and issue their own warnings, but that’s a slippery slope. The most you’ll ever hear me say is, “treat this LIKE a tornado warning” — usually that means I’m seeing something that I’m confident the NWS will also see, and will eventually post a warning for.) I bring all of that up because we are NOT obligated to show the SPC’s risk maps. We generally should — again, they are excellent at what they do — but if I have a different evaluation, I need to more-effectively communicate that disparity and why I’m thinking something different.
In my own defense, I also wrote this in the blog Wednesday morning:
Now, what ended up happening was that we warmed up enough, but the humidity didn’t increase enough — but I had at least outlined the source of uncertainty, and I thought I had communicated the LEVEL of uncertainty.
Because when you show one map that looks like weather Armageddon, and then follow it up with a discussion of how it might not be that bad, guess which one will get shared with over 100,000 people on Facebook? Guess which one people are going to remember and focus on?
No, I’m not calling you dumb. I’m saying its human nature. You’re not a meteorologist. In the same way the doctors, or lawyers, or experts in any profession sometimes forget that most of the people they talk to are at best peripherally familiar with the details of their field, meteorologists forget that most people think CAPE is what Batman wears. (It’s Convective Available Potential Energy, by the way.) Meteorologists think of instability in terms of a combination of temperature and humidity, while most people hear “instability” and think of a political discussion at Thanksgiving dinner. We have to get out of our own heads and make sure we’re looking at our own content from an outside perspective.
I wrapped up Wednesday’s discussion by saying what I thought that was a good, brief way of summarizing how to prepare:
The best advice is to hope for the best and plan for the worst…All of this means you should plan on staying weather-aware throughout the day.
Schools canceled for the day or let out early, businesses closed early, people canceled activities and dinners…WAY beyond what should have happened based on a forecast that still had significant uncertainties. We need to be more explicit: “don’t cancel anything until you check the forecast tomorrow,” or “have a plan B for outdoor activities, but the severe threat could still drop” or “for the love of God, just because you have snow days leftover doesn’t mean you’re REQUIRED to use them.”
Already Wednesday evening, it looked possible that thunderstorms along the coast of the Gulf of Mexico could throw a monkey wrench into the entire forecast. Before I went to bed Wednesday evening, I put this up on social media:
That might be the most-effective communication I did all week. When in doubt, remember the acronym KISS: Keep It Simple, Stupid.
I generally stand by what I said in Thursday morning’s on-air forecast and what I wrote in blog posts throughout the day. At that point, it was obvious that severe threat wasn’t as bad, but it still was unclear how much the severe threat would diminish. There’s a better balance there between “here’s what’s still possible” and “here’s why the worst-case scenario isn’t likely.”
The inevitable problem is that when we say “this could still change” or “this is still up in the air,” it opens us up to the lowest common denominator of internet comments (generally not spelled correctly):
– So you’re saying you don’t really know what’s going to happen
– Guess I’ll just look at the sky and figure it out for myself
– You guys never get it right anyway
– I’ll just watch Channel X, they at least made up their minds
That’s never going to change, unfortunately, and we just have to get used to it. (Frankly, I’m not at all worried about calling those people out here, because there is NO WAY those types of people will read all the way to the bottom of a 1400+ word article. I highly doubt they’ll even click on the link.)
To sum up…describing what went wrong yesterday from a weather-perspective takes one sentence: storms along the Gulf Coast blocked moisture from flowing up into the Midstate, so the humidity was too low for widespread severe weather. Hindsight is always 20-20, but from a purely analytical perspective, there’s not much data I could have looked at that would have had me ahead of the curve in playing down the severe weather forecast. The bigger problem was communication in advance of the threat, and that’s a trickier problem to solve — we have to make you aware of what could happen, but we also shouldn’t scare the tar out of you.
So we’ll start with this: what you’ll see in the future is not only an ongoing effort to improve our forecasts (an effort that never stops), but also a better effort to communicate “forecast confidence.” There’s nothing we can do about how people react to the information we provide, but we can do a better job in making sure that you’re fully armed to make the best decisions possible. And of course we have to remain accountable for our challenges or outright failures as well! We’ll do our best…