Published research tell us for every four tornado warnings issued in the U.S., only about one actual tornado touches down. During Friday’s severe weather outbreak in the Washington/Baltimore region, there were more than a few false alarms, prompting questions about whether the barrage of storms was overwarned and overhyped and, if so, what to do about it.
“It does seem like number of warnings is out of proportion with the actual storm damage we see in the DC area” commented Capital Weather Gang reader “stuckman”.
The concern about false alarms is that they will lead to public complacency.
“...if the weather forecasts overhype tornado warnings then when a REALLY SERIOUS situation happens, people may not pay attention,” commented DCRTVDave, whose public criticism of Friday’s storm coverage motivated this discussion.
Research backs the idea that false alarms can backfire. A 2009 study in the journal Weather, Climate and Society found “tornadoes that occur in an area with a higher false-alarm ratio kill and injure more people, everything else being constant.”
After the deadly Alabama tornado outbreak in 2011, broadcast meteorologist James Spann wrote: “I firmly believe apathy and complacency due to a high false alarm ratio over the years led to inaction in many cases that could have cost lives.”
In his book “When the Sirens were Silent” Mike Smith, senior vice president at AccuWeather Enterprise Solutions, documented that the NWS office in Springfield, Mo - which forecasts for Joplin, had a particularly high false alarm rate leading up to the tornado disaster on May 22, 2011 which killed 161 people.
The reverse argument (and reality) is that the NWS has a mission to protect life and property and is compelled to err on the side of caution in issuing warnings.
CWG reader Havoc737 put it this way: “Anyone who says these storms were “overhyped” [in D.C.] obviously doesn’t understand the potential for injury during severe weather - potential that increases exponentially if people are caught off guard.”
Following the the Joplin disaster, the National Weather Service post storm assessment noted those responsible for reporting and responding to severe weather are torn on the issue of how much warning is too much:
During interviews, Emergency Managers in particular felt the frequency of warnings was appropriate, while media staff were split with some saying ―most warnings were ‘cry wolf’”, while others emphasized the importance of advance warning for all tornadoes regardless of false alarms.
For Friday’s outbreak locally, the National Weather Service (NWS) office in Sterling, Virginia issued 23 tornado warnings for its forecast area covering the Washington and Baltimore region. So far, it has confirmed 10 tornadoes*. James Lee, meteorologist-in-charge of the Sterling office, said evidence for three potential additional tornadoes is being evaluated near Poolesville, Md., Manassas, Va., and Indian Head, Md.
A reasonable rough estimate of the false alarm rate for Friday’s outbreak was 50 percent - much better than the national average.
“I’m convinced we did very well [forecasting] this event,” Lee said.
Nevertheless, we have seen an outcry that warnings were “over the top.”
There’s no easy fix to the false alarm problem.
Ironically, improving radar technology may be increasing the potential for false alarms. Radar detection of the storms capable of producing tornadoes may be outpacing scientists’ ability to determine which of these storms will actually produce a real tornado on the ground.
Some have advocated we could reduce the perception of false alarms by changing warning terminology - perhaps by distinguishing “potential” tornadoes from “confirmed” tornadoes:
CWG reader Curt McCormick said: “I think NWS needs to use the term tornado emergency when there is a confirmed tornado on the ground and use tornado warning for doppler [radar] indicated storms. There needs to be a differentiator there.”
PhillyWeather added: “Tiered warnings (threat, warning, emergency or something like that) would really, really go a long way to undoing the “hype” factor.”
I agree with the gist of these comments. It’s long past time the NWS seriously evaluate and confront the false alarm problem. I’d just add new terminology would need to be carefully developed (with meteorologists) and rigorously tested by social scientists to assure effectiveness before making a change (from the current binary “warning” or “no warning” system). We don’t want to take one step forward and two steps back.
* Eight of the 10 confirmed tornadoes that occurred in Washington/Baltimore region on June 1 are listed on the NWS-Sterling website. Two more were confirmed Tuesday: small tornadoes that touched down near Simpsonville in Howard county and in Allegany county, Md. Additional tornadoes occurred outside the NWS-Sterling office’s jurisdiction in eastern Maryland and southeast Virginia.