Thursday, August 29, 2013

What Do Hurricane Katrina and the Moore Tornado Have in Common?

Hurricane Katrina. NOAA.
Yes, they were both "cyclones" (low pressure systems). Yes, they were both windstorms. But, there is something they have in common that might, if not managed properly, cause mass casualties in the future: Terrible traffic jams when a second storm days later.

Hurricanes Katrina and Rita

Many believe an unforecast major hurricane striking New Orleans followed by flooding the city would result in a five-figure death toll. When it comes to Katrina, people remember the seemingly shocking death toll and the flooding of New Orleans rather than the excellent forecasts, the successful evacuations, and the thousands of lives saved. The extensive media coverage of the plight of the Katrina survivors caused people in the region to "get religion" about hurricanes.

Less than a month later, major Hurricane Rita threatened the upper Texas coast, including Houston, and southwest Louisiana. It resulted in the third largest peacetime evacuation in world history -- about three million people! The issues with the evacuation were so severe that USA Today commented:

Hurricane Rita left this question: If a successful emergency evacuation involves 100-mile highway backups, motorists running out of gas and water, widespread road rage and the death of 23 seniors in a freak bus accident, what would a failure look like?

the newspaper went on to state,

If too few people evacuated for Hurricane Katrina, too many evacuated for Rita. Prompted by officials' broad, forceful orders and memories of how Katrina flooded New Orleans last month, nearly 3 million Texans fled on Wednesday, Thursday and Friday.

The difference was purely psychological: With the Katrina catastrophe fresh in peoples' minds, more evacuated than meteorologically necessary in Texas. Officials were caught off-guard.

Moore Tornado. Wikipedia
Moore and El Reno, Oklahoma, Tornadoes

On May 20, Moore was devastated by an F-5 tornado that killed 23. The tornado was televised live on television and the dramatic impact was heightened by two Oklahoma City television stations showing helicopter shots of devastated schools without knowing the fate of children and teachers inside.

Eleven days later, another F-5 tornado occurred south of El Reno -- about twenty miles from Moore -- in a sparsely populated area. All of the eight people killed were in vehicles. If that wasn't horrible  enough, literally thousands took to the highways from densely populated Yukon and south Oklahoma City just to the east of the El Reno storm. Why? In central Oklahoma, there are almost no underground shelters and media meteorologists were correctly explaining this was a violent tornado and people were safer underground (the same advice the NWS gives in those situations).

So, like in Rita, people both in and outside the areas threatened took to the highways clogging interstates 35, 40, 44, and 240. Officials were caught off-guard. Had the tornado continued east, those thousands would have been sitting ducks.

These incidents explain why social science is an increasingly important aspect of the storm warning problem. In all four cases, there were accurate storm warnings with plenty of advance notice. Yet, people made decisions based on emotions rather than the official advice.

So, I have some suggestions.

For Individuals

Storm warnings these days are quite accurate. I recommend taking the warnings and the corresponding advice seriously. There is no reason to believe the warnings for a second storm, a short interval after a major disaster, will be more or less accurate than usual. Do lot be ruled by emotions.

I do not recommend taking to the highway to escape a tornado threat unless it is a "particularly dangerous situation" tornado watch and you were willing to drive (perhaps for hours) out of the watch when it is issued and well before thunderstorms even form.

For Emergency Managers, Politicians, and Meteorologists

Too many politicians and even some meteorologists believe longer advance notice of storms (known as more "lead time") is some sort of holy grail. While I disagree, regardless of who is correct on this issue, if meteorologists give people an hour or more advance notice of a tornado in an area containing few or no shelters, residents will understandably try to get out of the storm's path. And, the numbers will be exaggerated if another major storm has occurred in the recent past.

Emergency managers work with politicians and meteorologists to create contingency plans for things we hope will never occur. It is probably time for EM's in both the tornado and hurricane belts to draw up plans that contemplate extraordinary numbers of people spontaneously deciding to get on the road.


  1. More "lead time"? We really get plenty of lead time on storm set ups days in advance already, and can start planning accordingly. True, the very specific tracks of a certain storm aren't available, say, three days in advance, but you know if the potential is going to be there.

    For example, last year we were planning on going to Lincoln, NE on a Saturday to pick up a new dog. The forecast for a tornadic Saturday was out several days in advance, so we were able to change our plans to Sunday. That Saturday we stayed around home, and missed being in storms such as the Geneseo-Langley-Salina tracked tornado. (We would probably have been coming back thru Salina at the wrong time!)

    The lead time to make plans already exists IMO.

    Mike B

  2. Mike, I agree. I believe the optimal tornado warning (not forecast) lead time is 15-20 min. before arrival.

    For most tornado situations information to the public is available days in advance. Watches are out hours in advance.

    I simply do not believe people will stay in a closet or cower in a bath tub for an hour or more. More warning lead time is likely to be counterproductive.


Note: Only a member of this blog may post a comment.