We witnessed a human tragedy this past week as Hurricane Dorian devastated parts of the Bahamas and caused damage to the coastal Carolinas. The well-being of human beings and their recovery should be the dominant headline, but it is actually a press release issued by the National Oceanic and Atmospheric Administration (NOAA). The Washington Post’s Capital Weather Gang headline read, “NOAA backs Trump on Alabama hurricane forecast, rebukes Weather Service for accurately contradicting him.” This stems from the flap over whether Hurricane Dorian was going to threaten Alabama and “Sharpie-gate.” Weather experts who consumed the probability maps correctly and understand how to interpret spaghetti plots knew that wasn’t going to happen (more on that later) as the week progressed. Models clearly latched on to a recurvature scenario up the coastline, and the National Hurricane Center forecasts reflected this fact as seen in the graphic below. Though claims about Alabama being impacted were propagating around social media and in Washington, the National Weather Service Birmingham office and prominent Alabama TV meteorologists like James Spann quickly refuted them. Amazingly, the drama dragged on for days and culminated Friday with NOAA issuing a statement that stunned many people, including former NOAA leadership.
Before I dive into the madness, let’s take a look at the facts. Meteorologist and University of Oklahoma doctoral candidate Sam Lillo tweeted the graphic above with the statement,
Because the gaslighting is getting real strong now, and it’s all too easy to start questioning reality, here is every forecast cone ever drawn by the NHC for #Hurricane #Dorian. Plus an arbitrarily-chosen state labeled
I think the graphic speaks for itself. However, here is where it gets weird. In a statement (see below), of which apparently nobody wanted their name attached too, NOAA wrote, “From Wednesday, August 28, through Monday, September 2, the information provided by NOAA and the National Hurricane Center to President Trump and the wider public demonstrated that tropical-storm-force winds from Hurricane Dorian could impact Alabama.” It went on to reference a series of hurricane advisories (#15 to #41) that supported this claim. NOAA, the parent agency of the National Weather Service (NWS), then rebuked its Birmingham office by saying in the release, “The Birmingham National Weather Service’s Sunday morning tweet spoke in absolute terms that were inconsistent with probabilities from the best forecast products available at the time.”
Elbert “Joe” Friday is a former director of the National Weather Service and a former President of the American Meteorological Society. He wrote publicly on his Facebook page:
The NHC and the WFOs (Weather Forecast Offices) are to be commended for the accuracy of the warnings and forecasts and their efforts to communicate the dangers to the emergency management community and to the public. The blemish to a Federal Government carrying out its duty to protect life and property was the erroneous warning issued by a Presidential Tweet. At the time of our glorious leader’s tweet, the chances of significant impact to the state of Alabama was essentially zero. The recent communications by a ‘NOAA Spokesman’ which tried to rewrite history in deplorable. Chastising WFO Birmingham for correctly pointing out that there was no danger to Alabama was unconscionable. NOAA is a great agency with an important mission for the nation. This rewriting history to satisfy an ego dimishes NOAA
The mission statement of the National Weather Service reads “provides weather, water, and climate data, forecasts and warnings for the protection of life and property and enhancement of the national economy.” This goal is significantly undermined if people start believing information on social media or doubting the expertise of meteorologists at the National Weather Service. In the graphic above, there are certainly very low probabilities during the timeframe mentioned in the press release that Alabama could experience tropical storm-level winds. Certain spaghetti plots also a had stray tracks going through that region. Yes, there were outlier or low probability scenarios for Alabama to be impacted, but that is where expertise comes in. We shouldn’t expect decision-makers to have to interpret information outside of disciplinary expertise. We have the meteorologists in the room for that. That probability map was produced days before assertions about an Alabama threat hit Twitter. By that time, the recurvature forecast was being disseminated by the National Hurricane Center.
In 2016, the National Weather Service New Orleans tweeted,
We know you’ve seen the spaghetti plots, but a little knowledge can cause a whole lot of trouble. Do you know which lines belong to which model and which ones are most reliable?
In a Forbes piece written in 2016, I urged caution in the use and interpretation of spaghetti plots because much of the public struggles to interpret complex probabilistic scenarios. I cannot tell you how many times people will zero in on outlier tracks or the low probability scenarios. It is also well-documented that many people don’t know what “percent chance of rain” actually means or how to interpret the hurricane “cone of uncertainty.” A potential lesson is that perhaps we need to rethink some of the probabilistic graphics we put out to the public or provide more context with them but that is a conversation for a different day.
Former National Hurricane Center Director Bill Read expressed his dismay over the NOAA press release in social media. He wrote on Facebook:
My understanding is the employees have been told they are not to comment on the issue. So I will. As I see it there are two possible drivers leading to the statement. Either NOAA Leadership truly agrees with what they posted or they were ordered to do it. If it is the former, the statement shows a lack of understanding of how to use probabilistic forecasts in conjunction with other forecast information. Embarrassing. If it is the latter, the statement shows a lack of courage on their part by not supporting the people in the field who are actually doing the work. Heartbreaking. I disagree with the essence of this statement. What the Birmingham NWS office sent out Sunday morning was correct and served the public well. It clearly let the public know that they were not at risk from the impacts of Hurricane Dorian. So disappointing.
There are good and smart people at NOAA. I have even rationalized a scenario in my mind that they are “falling on the sword” for the sake of their NWS office. While the optics look terrible, my gut tells me people within NOAA know that NWS Birmingham did the right thing. I sure hope so. If not, Admiral David Titley, the former Chief Operating Office of NOAA, may be on to something with his Tweet from Friday:
Perhaps the darkest day ever for @noaa leadership. Don’t know how they will ever look their workforce in the eye again. Moral cowardice.
Over the course of this ordeal, support for the National Weather Service-Birmingham has been consistent, and an #IStandWithNWSBirmingham movement has emerged. The National Weather Association (NWA), one of the two major weather-related professional societies in the United States, meets this weekend in Huntsville, Alabama. It should be a spirited meeting in light of this backdrop. Former NOAA Administrator Dr. Kathy Sullivan gets the last word. She wrote:
I told my Secretary (Commerce) and White House on several occasions that I would not “put my thumb on the scales” when I had satisfied myself that NOAA’s science was sound. She backed me, even when prior personal relationships and personal political capital were involved. I was prepared to resign over a couple of issues, rather than take a false stance, but we prevailed on evidence and logic in both cases.