BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

Three Problems With The Word 'Bust' During Real-Time Weather Threats

Following
This article is more than 4 years old.

Earlier this week the Great Plains experienced severe weather. This statement is not particularly unusual at this time of the year.  If you were following the majority of meteorological messages leading up to Monday, dire warnings about an epic tornado outbreak were being conveyed. In fact, NOAA's Storm Prediction Center (SPC) tweeted that morning:

The latest forecast from SPC has increased the tornado probabilities from 30% to 45% from northwest Texas into central Oklahoma. The last time a 45% tornado outlook was issued was during the Tornado Outbreak in Oklahoma and Kansas on 14 April 2012.

They also issued a Particularly Dangerous Situation (PDS) tornado watch for parts of Texas and Oklahoma. SPC noted that "This is only the second watch in SPC history where every category of watch probabilities (torn, wind, hail) are at greater than 95%." There was virtual certainty that these things were going to happen, and they actually did. On Monday May 20th, I personally watched tornado polygons illuminate my weather radar screen much of the day. Yet, the word forecast "bust" started creeping into the narrative of our insular meteorology community. Here are three dangers of the word "bust" in such real-time weather events.

NOAA

The timing of "bust" declarations. There was such a heightened sense of alert about the storms on Monday that many businesses, schools, and government agencies in Oklahoma closed. According to a report on CNN's website,

At least 19 tornadoes swept through central Oklahoma, Texas, Missouri and Arkansas Monday, carrying warnings of "considerable" damage to homes, businesses and vehicles with the possibility of "complete destruction."

Parts of Arkansas, Missouri, and Oklahoma were under tornado warnings and flooding threats early Tuesday morning as I write this article. With at least 19 tornadoes, flooding on Interstate 40, and hail damage, how could this be a forecast "bust?" A Tweet by Michelle Lynn strongly resonated with me:

For those calling it a bust, say that to my family who live in Mangum, OK. Luckily, they are ok, but that tornado was NOT A BUST...

I personally don' like the knee jerk "bust" declarations as a real-time event is unfolding. As a meteorologist, I certainly understand that when verification statistics are done, this event likely falls short of some of the messaging metrics associated with "outbreaks."  From a meteorological perspective, was it a bust? If you are using the metrics of "45% coverage of tornadoes or long-track, violent tornadoes," then the Monday outbreak probably doesn't meet that threshold. However from a human toll perspective, I say "So what?" The messaging forced people to act and hazardous severe weather (and flooding often overlooked) happened. My concern is that our weather geekery and verification statistics don't really matter to someone that was impacted. I always default to the human responses before meteorological numbers. I think there is a time for the discussion about forecast verification, and it is not in the midst of an evolving, dangerous event. As "bust" was making it around social media, tornadoes, hail, and intense rainfall were still on the table for many people, and they needed to remain alert.

Hype or hope? Such events reveal a certain scientific giddiness coexisting with a sense of dread among meteorologists and weather enthusiasts. Many of meteorologists, including me, pursued weather because of hurricanes, storms or awe-inspiring observations. A rare PDS outbreak stokes our basic curiosities even as we know the storms could alter lives in an instance. However, it is important to realize that storms are not entertainment or games.

From a messaging standpoint, many people already feel that "weather" is hyped. The era of 24/7 news cycles, extreme chaser videos and social media are partly responsible. I get it. My Forbes colleague Dennis Mersereau's tweet is spot on:

Today, and I cannot stress this enough, is why you don't hype the weather. Once you're using EXTREME and VIOLENT with regularity, where do you go from there when the red lights start flashing?

His point is that overuse of certain words can desensitize the public to events like yesterday. Dedicated meteorologists like those in the picture below recognized that "level of alert" for this event was warranted. The meteorological ingredients (instability, wind shear, lifting mechanisms) were certainly in place to produce numerous tornadoes, flooding, and hail. However, "outbreak forecasts" can be challenging too.  I previously discussed "pros" and "cons" of long-lead time outbreak forecasts in Forbes. Early indications suggested that an area of stability (a "cap") may have somewhat suppressed the potential powder keg. Ironically, caps can be key ingredients in explosive storm development too.

NWS Norman

The main concern for me is that we have become a culture that frowns upon preparing for the worst and nothing happens. This has always been somewhat baffling to me. At times, there seems to be a disappointment factor if nothing happens, and there was preparation. I call this "worst-case scenario bias." By the way, a few weeks ago my family had to rush to the basement because of a Doppler radar-indicated tornado. We ended up being late for an event, but I would do it all over again hoping for the best outcome. We buy car or home insurance with the intent of never having to use it.

The realities of false alarms. Having said all of that, the reality is that false alarms are a challenge in weather messaging (or are they?) A new study entitled "Cry wolf effect? Evaluating the impact of false alarms on public responses to tornado alerts in the Southeastern United States" was just published in the journal, Weather, Climate and Society. They reported that roughly 75% of tornado warnings in the U.S. are false alarms. Such high rates could lead to a "cry wolf" situation in which people change how they respond to future warnings. In the study, 4162 residents of the southeastern U.S. were interviewed. Surprisingly, the researchers found that concerns about false alarms generating a complacent public may be somewhat exaggerated. A 2015 study in the journal Risk Analysis, "The Cry Wolf Effect and Weather‐Related Decision Making," also found conflicting results. The abstract  says that "very high and very low false alarm rates led to inferior decision making, but that lowering the false alarm rate slightly did not significantly affect compliance or decision quality." These results are counter to previous studies that linked "cry wolf" with non-compliance to warnings. There are many factors that likely influence a decision to act on a warning: optimism bias (it won't be as bad as they say), the psychological evaluation of risks vs. cost, geography or lack of trust of the warnings themselves.

These challenges highlight why the emergence of social science research on communication, psychology, sociology, and geography is so important. Our weather forecasts continue to improve as our technology and understand advance but tornado forecasting is still hard. We need continued integration of technical and social science research as we have seen recently in the VORTEX-SE campaign.

Follow me on TwitterCheck out my website