Certain Death Warnings

[Patrick Marsh]

If a 20 to 25 foot storm surge into Galveston isn't the time to pull out the OMFG language, when is?

Then what do you say with a 30-35 foot surge or greater?

 

[Ryan McGinnis]

Then what do you say with a 30-35 foot surge or greater?

Does it matter? That's like asking whether you'd like to fall from a 50 story building or a 75 story building. The results are the same. The warnings should be equally dire. Galveston can take a 10 ft surge. Galveston might be able to take a 15 foot surge. Anything 20 and over and chance of survival plummets. "You goin' die!" language is very much appropriate at that point, IMO.

 

[rdale]

But are you guaranteeing it will be 20-25ft? What if it's 15-20ft? I'd call that a possibility, but using "you're going to die" language means no matter what - if Ike ends up veering at the last minute and the surge is only 10 feet, you're going to die?

 

[Ryan McGinnis]

But are you guaranteeing it will be 20-25ft? What if it's 15-20ft? I'd call that a possibility, but using "you're going to die" language means no matter what - if Ike ends up veering at the last minute and the surge is only 10 feet, you're going to die?

That decision is made by professional people with more information than you or me who have a very strong grasp of probabilities and risk assessment. If the NHC says "get out or you're going to die", they aren't just yanking on people's chains. At what point does your logic stop? Should we not be issuing tornado warnings for doppler indicated rotation, because only a small percentage of doppler indicated tornadoes actually produce? Should we not issue flash flood warnings until someone actually goes out and measures all the local rivers and streams?

You work with the data you have and an understanding of statistical probabilites. The argument that people seem to be making here is that the statistical probabilites of death in storm surge are just not high enough to include language of death. To you I would ask: under what statistical analysis/model do you base your assertion that the warning issued by the NHC citing "possible certain death" is overblown? Do YOU have access to realtime SLOSH models and volumes of pre-run FEMA surge simulation studies? Or are you maybe assuming you know a little more than you know?

 

[rdale]

I'm just saying - if a large part of those 40% staying behind live, what do you do next time?

(This wasn't NHC using the "die" line, it was NWS Houston)

 

[Patrick Marsh]

That decision is made by professional people with more information than you or me who have a very strong grasp of probabilities and risk assessment. If the NHC says "get out or you're going to die", they aren't just yanking on people's chains.

Please point to a product in which the NHC says "get out or you're going to die." I've only seen these products from local WFOs.

At what point does your logic stop? Should we not be issuing tornado warnings for doppler indicated rotation, because only a small percentage of doppler indicated tornadoes actually produce? Should we not issue flash flood warnings until someone actually goes out and measures all the local rivers and streams?

Please point to a product that you have listed above that has "[take xxx action] or you're going to die". One could argue that Tornado Emergencies fall along these lines and I think my opinion on these are well known on this board.

You work with the data you have and an understanding of statistical probabilites. The argument that people seem to be making here is that the statistical probabilites of death in storm surge are just not high enough to include language of death. To you I would ask: under what statistical analysis/model do you base your assertion that the warning issued by the NHC citing "possible certain death" is overblown?

Unless you include the actual probabilities, I don't think the death language should ever be used. We aren't as good as we want people to believe. You lose all credibility the first time your death situation doesn't pan out.

I'll let slide how "possible certain death" is a complete oxymoron...

Do YOU have access to realtime SLOSH models and volumes of pre-run FEMA surge simulation studies

No, I don't have access to either of these, but I agree with what rdale is saying. You should NEVER use scare tactics because if it isn't as bad as you say it is you are going to lose their trust. Just look at how many NOLA residents say they won't evacuate next time...

 

[Mike Peregrine]

Looks like the original language was taken slightly out of context by the media (surprise surprise). They took: ' residents living in single story dwellings may face certain death' and turned it into:

CERTAIN DEATH 2008
Stay tuned to blah blah blah and watch as peoples' lives are lost one by one.

Shameful.

It's true, though. I don't think it will mean certain death for all those people. And then next time ... on the Cat 4 ... that's a hundred times the risk ... and everyone is saying "you'll die, you'll die" ... everyone will laugh and dress up like bears. Only this time they really will die.

 

[Chris Nuttall]

I mean no offense to anyone here. This applies to all of us, including me...especially me. It's easy to criticize and point and say, "That's stupid. You shouldn't do that for xxxx reasons." But truthfully, how would any of us handle this situation if we were on the hot seat? I don't have the experience yet to say what I would do. It's very easy to scrutinize from the outside.

There's another thing to consider, though. WFOs work very closely with local and state emergency management agencies. Maybe an EM requested that some enhanced wording be used.

I've been in a WFO during tornado events. There's typically a lot of things going on at once...especially when it's right outside your window. I can only imagine how tough it must be to work during a landfalling hurricane.

 

[rdale]

That's why we professionally discuss things -- that's the StormTrack way

There are EM's here, there are met's here, there are probably people who study risk management / psychology here. Nobody is calling anybody stupid. I think it's healthy for people to address potential issues for future use.

And note - Houston dropped that wording pretty quickly today.

 

[Mike Peregrine]

I can't relate to what it must be like to be in that position (and wouldn't want to), but at the same time can't help but wonder at the situation we've created for ourselves. The way that people respond to warnings is an interesting subject. Lots of thinking processes at work. There is also the question of rational vs. irrational fear, and which of these is the healthiest way to achieve desired results. During Gustav, we heard Mayor Nagin make some pretty incredible 'storm of the century' remarks, and then this was followed by the 'certain death' scenario yesterday. Without boring people with the details of my opinion here, I blogged a couple of articles on the subject between last night and this afternoon. I feel like it is worth taking a look at what may be behind the reasons people make the choices they do.

 

[rdale]

Good summary Mike...

NWS said that if you stayed, it was almost certain you would die. 140,000 people stayed behind.

What will those 140,000 people think next time? Or what about the million+ people that left, and see those who left behind still living when they come back?

 

[Dennis Sherrod]

Good summary Mike...

NWS said that if you stayed, it was almost certain you would die. 140,000 people stayed behind.
What will those 140,000 people think next time? Or what about the million+ people that left, and see those who left behind still living when they come back?

We can probably figure that out from your statement. If they die, they won't have much to think.

 

[rdale]

I don't think they all died though... So what do the survivors think next time they are told to evacuate -- not because of problems they'll have afterwards, or the potential destruction, but "because it is certain you will die if you stay"?

 

[Ryan McGinnis]

Good summary Mike...

NWS said that if you stayed, it was almost certain you would die. 140,000 people stayed behind.

What will those 140,000 people think next time? Or what about the million+ people that left, and see those who left behind still living when they come back?

A massive number of people didn't leave after being told point blank they were quite possibly going to die if they didn't. THOSE PEOPLE WOULD NEVER LEAVE FOR ANYTHING so it doesn't matter what they heard in the warnings this time. They'd stick around for a Godzilla attack. They're going to be just as retarded next time as they were this time.

My eyes are rolling so hard I'm pretty sure they can hear it all the way in Texas. You're ignoring a large part of the equation: what would have happened if the much larger number of people HADN'T evacuated? I am not trying to build an argument that the NHC was certainly right to include that language (I don't have the data), but I think that the argument you keep making is rather silly and a touch arrogant without some sort of hard statistical data. The weather service is not Accuweather, it's not a politician -- its sole mission is to provide the public reliable information about what the weather is going to do and inform them of dangers. It doesn't hype things up for ratings, it isn't trying to save its elected position, it's trying to save lives. They've done a bang-up job in the past getting life-saving information to the public, so if you're going to say they overdid it this time, you'd better have some statistical models at hand that clearly illustrate that your position -- that people should not have been warned that there was a strong threat to their lives -- was not only right, but demonstrably right at the time that the NHC put out their advisory. Otherwise you're armchair quarterbacking an incredibly intricate and complicated job in the very worst way.

Personally, I think that not telling it to the people straight for fear that if it doesn't verify nobody will take you seriously "next time" is silly. If you take this argument out to it's inevitable conclusion, we should not be issuing tornado warnings until we see an actual tornado on the actual ground, and we should only issue them for communities we are 100% certain are going to be destroyed. Otherwise, next time, people won't heed a tornado warning. "Ah, yes, that half mile wide long track tornado is 10 miles from Oklahoma City and headed right for it but, you never know, it might lift at the last minute... better lay off the warning for now. Otherwise OKC won't take us seriously next time..."

 

[rdale]

Tornado Warnings don't say "you will certainly die" in them. Your comparison is not one I (or anyone else here) will be making.

And to clarify - NHC did not use this language, NWS Houston did. Only for a short time, then they removed it shortly afterwards.

You lose credibility when you say "140,000 people will most certainly die" and they live. Will people believe Mayor Nagin when the next one is "the mother of all storms"? I don't think you can have hard data for this sort of equation. Common sense does have some value still in this field.