Wednesday, March 25, 2015

Longer Tornado Warning Lead Times? BAD Idea!

An Oklahoma representative wants to pressure the National Weather Service to issue tornado warnings one hour in advance. That is a terrible idea.

Multiple, convincing studies demonstrate that 15 minutes is the optimal lead time for a tornado warning. Why?
  • People will only stay in basements, bathtubs and closets for so long.
  • More than 18 minutes, the death rate goes up. This is likely due to the warning losing its urgency.
  • As we saw, twice, in central Oklahoma in 2013, people without basements will hit the highways, likely putting themselves in more danger with they have more notice. 
  • The false alarm rate will go up, possibly way up.
The death rate (deaths per million population) has been cut 97% by the tornado warning system. A Nobel Prize-worthy achievement. Unfortunately, in an attempt to get at that last 3%, a lot of bad ideas  have been implemented and proposed. 

Congress should vote this one down. 

2 comments:

  1. Hi Mike, I agree with your assessment.

    One question.

    How does a product like what you showed recently (long term - ie hours prior - tornado prediction) compare in your mind with a one-hour-away tornado warning?

    The latter seems like a terrible idea, as you have noted, because 5 minutes can feel like an eternity in a basement, let alone an hour. But could a several-hours-off tornado prediction (should it prove to be as accurate as was shown) do the same?

    ReplyDelete
  2. Good question. The difference is in the way the product is going to be used.

    Forecasts based on our new computer model will go to our commercial clients only. For example, an experimental forecast made at 2pm Wednesday had Moore in an "enhanced tornado risk" area when the NWS had only a severe thunderstorm watch in effect.

    Think of it as a more precise tornado watch.

    ReplyDelete