"Certain Death" Revisited - Long post

[Scott Overpeck]

Quoted from thread on "Will Ike be retired" in the hurricane discussion forum.

Because 150,000 people stayed in the evacuation area, and were told they would "certainly" die. Yet over 99.9% of them lived, at least as far as I've seen (I assume if the death toll exceeded 100,000 we would know by now.)

So what happens the next time NWS says "evacuate or you will certainly die"? I say it's better to tell people what conditions they'll face, what it will look like after the disaster is over, remind them of Katrina, etc. - not "you will die if you don't leave."

Fool someone once... odds are it won't happen twice...

I have responded to this here in a new thread because it is a long response and will probably generate more discussion. My goal here is to provide a different perspective and if moderators feel it is not appropriate, please remove it or do with it as you please.

It has been one month since Ike hit Galveston. I have tried really hard to stay away from this topic since I am a forecaster in the very office that issued the HLS with the certain death wording. I even re-issued a couple of HLS’s with this wording in them because given the data and facts we had at the time, it seemed very appropriate. I have had some time to think about the events during Ike. With that said, I have stayed away because I realize it is much easier to criticize from afar, anonymously, and do so without all the facts of the situation.

While I do not feel obligated to answer critics, here is more detailed information that we based the decision to use certain death wordings in our HLS. The evacuation area we issued the certain death call to action (CTA) was for Galveston Island and Bolivar Peninsula. It just so happens that Chambers, Harris, Galveston, Brazoria, Matagorda and Jackson Counties are included in NHC’s hurricane warning which is the basis for the HLS. We cannot separate counties from this warning as it is not our local office’s warning. We can however segment out other counties for hurricane wind warnings, etc. The certain death call to action was placed in the storm surge part of the HLS because of a number of reasons.

First let’s look at some meteorological data for some background. The first HLS with the certain death CTA went out with the 15z Thu NHC advisory – advisory 42. The track from Wednesday to Thursday has shifted more to the north towards Galveston which increased the storm surge for Galveston. By this time, Ike was forecast to make landfall between Freeport and Galveston. This was also when the hurricane warning was issued for the area. Ike was a strong Cat 2 for wind intensity, but had a large wind field. This was a unique storm, different from a lot of hurricanes with this aspect. The large circulation center did not end up contracting, but the thinking was still for Ike to become a Cat 3. So, 36-48hrs out, we are still thinking of the possibility of a major hurricane. Not only that, but the storm surge data was coming in well within surge values associated with Cat 3/4 storms not Cat 2. SLOSH had storm surge of 15-17 feet for Galveston, 17-20 feet for Bolivar, 20-22 feet in the Bay, and as much as 23 feet in Trinity bay and Houston Ship Channel. Some of the later SLOSH runs had similar values but there were a few that were higher and a few with lower values. It really depended on the track of the storm and the intensity. A track shift of 20 miles made a significant change in storm surge. Even if the track is right, there can be as much as a 20% error in surge values from SLOSH. So, at the time, we did our best to convey what to expect but could not pinpoint more exact surge values for a given location. Our HLS at 15z Thu highlighted storm surge of 12-16 feet from Matagorda to High Island with 15-20 feet possible in both bays. The range of 15-25 feet in storm surge actually did occur in later HLS to account for error in the track and to convey the higher end of the threat. It also applied to Galveston Bay. Yes it is a big range, but breaking it down would not properly convey the threats given the possible errors in the track at the time Thursday. By Friday when the track was becoming more evident, we were able to use SLOSH to narrow down surge values for specific areas. We knew that surge of 15-20 feet was more likely, but at this time, what if the storm strengthened to a Cat 3? Wouldn’t 25 feet be possible just as much? What if the storm shifts at the last moment (which it did, and was a big reason why the surge was not more than 20 feet in some areas, but it very well could have shifted so that the surge was more)? The point being that Thursday morning, given the uncertainty in the forecast track and any track errors, given the surge values SLOSH was predicting, it made sense to use the certain death CTA to get people’s attention to this threat. People were not taking the threat seriously.

Now let’s add some situational awareness to the equation. At Thursday morning, Galveston Co was doing mandatory evacuations for the west end of Galveston and surge prone areas along Galveston bay like Kemah and San Leon. The people in these areas were not taking the evacuation seriously and people were slow to make any preparations. The Galveston Co OEM kept telling us that people were not getting out that needed to get out of surge areas. The people were not taking the threat of surge seriously. Galveston Co ended up issuing mandatory evacuations for all of Galveston starting noon Thursday which was after we issued the certain death CTA with the hurricane warning. People started to evacuate by this point. Given the meteorology and the fact that people were not evacuating, we continued the certain death CTA. People needed to get going and get out of surge prone areas. Let’s also add to the fact that this area had not had a major hurricane since Alicia in 1983. People here have no reference to what 15 feet of water will do to their houses if you only live at 4 feet above sea level. We have a major hurricane awareness conference every year in Houston, and do many preparedness talks every year, but people do not care until it is about to happen and by then it is too late. It had not happened to them for 25 years, a whole generation of people that have no clue about what a hurricane like Ike will do. You can remind them of Katrina all you want, but people here remember how Rita missed this area and they were stuck in traffic for 18 hours. At this point, as much as you tell people to be prepared ahead of time, it does no good when all they know is a traffic jam during the Rita evacuation. So, that leaves us with putting a seriously grim call to action to get people to understand the situation. Most of the general public have no reference for storm surge or any situational awareness to tell them to get out if 17 ft of water is coming for them. If you live on Bolivar where you are 3-5ft above sea level, and there is 17 feet of water not to mention wave action on top of that coming for you, you need to get off of Bolivar. No matter how high your house is built or how well, 12 feet of water will remove your house off its foundation. This also does not account for the possibility of even higher surge. If you stay in your house in that situation, I’m sorry, you are going to face death. Not only that, even if you do survive the surge, will you survive being stuck on a deserted island with no drinkable water, food, utilities, or medical services? Will you be able to fight off any alligators after being taken in the surge to Chambers County? To summarize, GCOEM was telling us people were not leaving, so we needed to do something to get their attention. If you live in a storm surge area, you need to leave, or else you take your life into your own hands and you may not live.

As it happened, the storm fortunately turn more to the north about 6-12 hrs before landfall with a nice wobble. Still, with that, you had storm surge of 12-14 ft for Galveston Is, about 12 feet for Galveston Bay areas like Kemah and San Lean and even smatterings of 14-16 feet measurements. Some areas in the Houston Ship channel at 12 feet with isolated measurements of 14 feet. Supposedly there was as much as 20 feet in mouth of Trinity River. There was 15-17 ft surge for Bolivar and 17 feet across Chamber Co. Surge had pushed 15-17 miles inland across Chambers Co. These are all unofficial and preliminary surge numbers, but I’d figured I’d share a little here to give you the idea that even though we had 15-25 feet in our HLS, our low end number of 15 feet was well in range, and our 12-16 feet for Galveston was well on target. Our low end number was certainly on target for Bolivar, and looking at damage photos, it will be a while before Bolivar will recover if at all. So, yes, now people have an idea of what storm surge can do. I still think that given the meteorological data and the situation that people were not evacuating, we needed to use a CTA to get people’s attention that they need to evacuate. Your life is at stake. So yes, the death toll is not high, but you still have 300-350 people missing. You still have crews finding bodies in debris fields of Chambers Co. The fact is that some of these missing people will not be found. Even with the CTA, there were 40,000 people that did not evacuate from Galveston Is and about 400-500 people that did not evacuate from Bolivar (based off media and GCEOC). Some of these people required high water rescues on Friday, and by Friday afternoon, conditions became too dangerous to get the rest of them out. I have a huge respect for the emergency personnel that participated in the numerous rescues in dangerous conditions. Even by Friday morning you had most of Bolivar under water with flooded roadways due to surge.

I’m not sure where you get 150000 people not evacuating. Galveston Island may have 60,000 people on it total, and about 280,000 total population for the county. So, ya maybe 150,000 people of the whole county did not evacuate, but not all of Galveston Co was under a mandatory evacuation. It was just the storm surge areas under mandatory evacuation, but the rest of the inland area of the county was under voluntary evacuation. Most people along the Bay evacuated, it was mainly the folks on the island that did not leave. So, yes 40,000 people survived and did not face certain death. In that aspect I am glad they lived. I would have hated to see that many people dead. I would not be able to sleep at night knowing that 40,000 people died because I did not do everything possible to warn them. So yes, from that point of view, the certain death wording failed and I’m glad it did. But by your logic, if 40,000 people died, then the certain death wording would have succeeded. To me that is a grave failure because now you have 40,000 people dead. At the same time, I think the certain death wording did what it was designed to do. It got the attention of people to take this situation seriously and hopefully got people to act where as before they were not going to act. I hope that hearing the possibility of death people said, “You know, I think I need to evacuate.†The sad thing is that you still had 400 people left on Bolivar and I imagine a good chunk of those make up the 350 or so that are still missing.

Finally, I ask you, what would happen if we had not used such wording like “certain death†and there was 20-25ft of surge that wiped out thousands of people who did not evacuate? People would crucify the NWS because we did not warn them of this possibility when we had the data to support it. Ultimately, there would be a whole new staff at the NWS HGX office because of a Congressional investigation showing we did not do our job. Personally, I would have a hard time living knowing that I could have used better warnings to better convey the threat from the surge and save lives. I would much rather deal with people criticizing us for using certain death with people surviving than the alternative of not using such wording and having a bunch of people dead.

I have appreciated everyone’s candid opinion on this matter as I want to understand how our warnings impact people. I want to know that we are making a difference with what we do. Your thoughts have given me a good insight into how people respond to our warnings. Thank you for reading this long rant, and I apologize if I offend anyone with it. Again these are my views on the topic, not who I work for.

 

[rdale]

I appreciate the insight...

"Not only that, even if you do survive the surge, will you survive being stuck on a deserted island with no drinkable water, food, utilities, or medical services? Will you be able to fight off any alligators after being taken in the surge to Chambers County? If you live in a storm surge area, you need to leave, or else you take your life into your own hands and you may not live."

I think those are pretty powerful statements... Using them in the HLS would likely get the point across. Would it convince as many people to leave as "certain death" would? I'm not sure. That's why I'd like to see what research results come from this event.

In any case, the 40000 people (160000 came from press reports of the number of people who didn't leave mandatory evac zones but I'll take your word) who survived "certain death" have a choice to make for the NEXT event. If no hurricane hits the area for 30-50 years, the CD call was fine.

But let's say that next year another Ike comes ashore. In the same area. How are you going to convince the "certain death" survivors that "we really, really mean you are going to die this time"? Had the HLS terminology been your above comments about gators and floodwaters and no rescue and the like and not CD, you could say "like last time, you'll have to fight alligators and we won't come get you." I think that would get the point across.

I'll relate to Tornado Warnings... Even in Norman's Tornado Emergency, with a confirmed large tornado on the ground, the phrasing didn't change to "If you are in the path of this storm, you face certain death" but I think the TE did the job. (Visit past threads on "will TE's that bust lead to degraded usefulness in the future?")

So to sum up my shorter rant -- in a forecast product, tell users what you think will happen and let them make the decision. I think a detailed listing of post-Ike conditions as you listed above would convince people to leave. And if they didn't leave - you did your best. It's up to EM to get people trained in advance, communicate that threat ahead of an event, and emphasize the risk to their community.

"CD" may have done the job, and nobody is questioning that. My concern -- what happens ahead of the next storm with the large number of people who are still alive after forecasts said they would be dead.

- Rob

 

[Greg Higgins]

I would like to add a different viewpoint but along the same lines as Scot and rdale. We have had several large events in the past several years where loss of life and property have been horrendous. The problem that I see is a political challenge, as in the aftermath the affected population blames the NWS (or other government agencies) for either not warning them or for trying to scare the s*** out of them. The fallout, as Scott pointed out, is a government oversight investigation that eventually lays the blame (and the loss of their job) at the lowest level government employee that was just trying to do their job, usually with little lead time to get the message out (speaking of short term events such as tornado's).

As someone in the emergency service field for 30 years, I have become cynical and have embraced the saying, "people are stupid". As an occupation, we are taught from the very beginning (be it EMS, fire service or law enforcement) the politically correct terminology of attempting to educate the public of hazards and never laying blame or pointing fingers when something bad happens. Our country has changed over the decades since WW II and become "soft", not wanting to accept responsibility for their actions and expecting the government to step in and take care of them for the lack of their own inaction. In some shape, form, or fashion, this must come to an end!

The Houston Fire Department at one time (not sure if it is still being done) would bill for their hazardous material response. This included the cost of the equipment, manpower, etc. to recoup the cost to the city for someones "mistake". Colorado and several individual Search and Rescue teams routinely bill the victim (if found alive) or their family for time and equipment spent for the search and rescue activities. A good recent example is Steve Fossett. It may not seem like much, but billing a person for their stupidity hits them in the wallet and seems to make a difference! Lawsuits to deny these charges have been filed and struck down numerous times based on "what a reasonable person would have done".

It amazes me that people will risk the lives of others (the rescue team) to save them when their inaction places them in grave danger. The mayor of Galveston (may be wrong on who this was) really gained my respect when he put out the warning for the city and told folks to get out and if you choose to stay, there would not be any rescuers until it was safe for them to venture out. Just remember, if we have to respond in the height of the storm to rescue you and we are injured or killed, have our equipment damaged or destroyed, who is going to rescue us and then have to rescue you??? And then we get sued for dying in the attempt and not providing service to you! What a screwed up thought process. At some point, people MUST take responsibility for their own actions!

In that vein, I agree with Scott that something must be done to encourage the public to understand how dangerous a situation may become and that extreme wording should be used, but as rdale pointed out, when you use the term "certain death" and nothing happens, those folks will now turn around and sue (as well as have a negative memory of the issuing agency, and fail to respond at all the next time something occurs because they can't "trust" what is being said) for overstating the danger! Use strong wording as rdale suggested of the consequences that may occur, do it often, and in several different ways. And get the local government officials to grow some balls and back the emergency services groups. They will be more than willing to jump in, even in the face of extreme danger to themselves, to serve the public, but there is always a line.

I hope this discussion at some time will bear fruit and all parties will be able to make better choices in the future.

Greg Higgins

 

[Shane Adams]

I've read these posts and my opinion remains the same: people are responsible, at the end of the day, for their own actions. A little thing called "personal accountability" applies here. You have a choice, and if you choose to stay, you get what you get. We've been coddling ignorant Americans in natural disaster situations long enough. You shouldn't have to be brilliant to understand what can kill you. The problem is the feeble human mind can't fathom the reality of something happening to them when they can barely wrap their mind around the idea; "there's no way a few waves can do any harm"..."it's just wind and rain". Ad to that denial ("It won't ever happen to me") and a dash of complacency, and there you go.

It's not anyone else's job to fend for the lethargic. If preserving your own life isn't enough motivation for you to act, then why the hell should anyone else?

 

[Steve Miller TX]

Finally, I ask you, what would happen if we had not used such wording like “certain death” and there was 20-25ft of surge that wiped out thousands of people who did not evacuate? People would crucify the NWS because we did not warn them of this possibility when we had the data to support it. Ultimately, there would be a whole new staff at the NWS HGX office because of a Congressional investigation showing we did not do our job. Personally, I would have a hard time living knowing that I could have used better warnings to better convey the threat from the surge and save lives. I would much rather deal with people criticizing us for using certain death with people surviving than the alternative of not using such wording and having a bunch of people dead.

This is often the very thing that the "critics" overlook...that and the fact they aren't responsible for providing warnings in such a dire situation. It's amazing how people's tones and attitudes change when it's their arses on the line. Quoted from Scott: "To summarize, GCOEM was telling us people were not leaving, so we needed to do something to get their attention". My question to the critics: What would YOU have done?

How easy it is to critically analyze the NWS actions AFTER it's all said and done. As you pointed out Scott, a wobble to the left south of Galveston Island would have certainly produced more sinister results. Nobody could predict with certainty the exact path Ike would take far enough in advance for people to make decisions. You go with the envelope of potential error in such forecasts and warn accordingly. The old saying goes that if you plan for the worst and hope for the best, you'll not be surprised when the worst happens. Please refer to Scott's quote above.

It's clear now that "certain death" unfortunately prevailed for residents that remained on Bolivar Peninsula. When I looked at all of the arial views of the aftermath, I found it horrifying. I also said to myself, "The NWS guys were right". I don't see how anybody living anywhere along the coast could reasonably have a heightened level of complacency regarding the warning language used in this case. Of course, as Shane pointed out, there are ignorant people out there who will. Complacency and ignorance are the true dangers to people. Nothing in the world will ever change that short of a near-death experience. Certainly then, it is ridiculous to believe that language in a particular warning product would ever have any bearing on that. Personal accountability cannot and should not be expected to be controlled or mandated by the government. Again, using Shane's quote: "If preserving your own life isn't enough motivation for you to act, then why the hell should anyone else?"

Indeed.

So Scott, I fully support your and the NWS decision to issue such a dire warning to residents down there. I'm sure many more people heeded it than ignored it. Ya'll did an excellent job in the face of such a HUGE impending disaster. My hat is off to you guys. Thanks for the very detailed and thoughtful post.

 

[Jim Saueressig]

People will respond poorly with the next "possibly deadly" event because not enough people died during Ike. Personally I think the term "certain death" was a mistake based on the way people will react to such "failed" statements during the next event. I understand the many reasons terms like that are used which range from protecting people to eliminating the chance you will be sued because you didn't warn people enough.

Maybe next time a term like "High death toll potential" or something related to a "potential" for an event to occur rather than a "certainty" that it will occur would be more appropriate.

People are stupid in many ways and quite rebellious in others so a statement to the "potential" threat will get the point across to the stupid ones while not giving the rebellious ones any reason to rebel....

 

[Joe Dorn]

People will respond poorly with the next "possibly deadly" event because not enough people died during Ike.

I am very active in the Bell County emergency response groups, usually in the County EOC. We are the receiving county for Brazoria County (just west of Galveston) evacuees. One reason the death toll was low was because our shelters were full.

The State of Texas DEM has done a fantastic job of preparing for hurricanes and the aftermath since Rita/Katrina. The evacuations were not publicized that much because they worked so well. There were glitches but over all it was a top notch job.

Instead of hanging around the Super Dome moaning and crying, the folks were setting in the shade playing dominoes, enjoying three well prepared meals each day and sleeping in a clean bed. We were told by the Emergency Manager of Brazoria County that a lot of them wanted to stay with us.

As far as how deadly it was, ask those people that chose not to evacuate and ended up in the ruble piles on the north shores of Galveston Bay or beyond the third sand bar in the Gulf of Mexico.

Yeah, it feels good to be part of the solution...

 

[AlDanks]

In the case of hurricane landfall, I wonder if it is feasible to produce a % likelihood of death map (or the inverse % likelihood survivability).

If one is contemplating serious surgery one typically gets % success, % survival. It would seem that people in the areas with high % death (low % survivable) would leave given the forecast.

Essentially the "certain death" forecast a 100% death, 0% survivable. Perhaps reality was closer to 90%, 10%. How many people would have left with the 90/10 forecast? Probably most that left with the certain death forecast. Especially if an additional warning that changes in track or strength of the storm may further reduce likelihood of survival.

I think a 90/10 forecast would not be treated as a 'false alarm" like the certain death might be in future storms. Yes, people need to evacuate, but not just for the storm at hand - they need to evacuate for future storms too. The "certain death" forecast may have moved people for this storm, but possibly at the cost of making it more difficult to move some people for the next one. A 90/10 forecast might not have had such future cost.

I'm not trying to tear down what the forecasters did do, just offer a suggestion of a possible improvement. I think the forecasters did the best thing they could do with the system that was in place. Their forecast probably saved hundreds of people. I bet that no one who stayed and managed to survive would quarrel too much with the "certain death" forecast they were given. I bet they would agree that the forecast was accurate and they were extremely fortunate to still be among us.

 

[Joey Ketcham]

I have read people saying that the "certain death' statement failed, I'm confused by this. The whole goal of that is to get people out and prevent high death rates, so if there are minimal deaths after a certain death statement is made then doesn't that mean that it was a success?

Saying it failed, to me, would mean that a certain death statement was made and you still had thousands of people die. Maybe I'm just reading it wrong?

 

[rdale]

You'd probably need to post some examples or talk to those who've made the "failed" comment, I've not seen that...

 

[Joey Ketcham]

You'd probably need to post some examples or talk to those who've made the "failed" comment, I've not seen that...

Well here is one posted right here in this very thread:

Personally I think the term "certain death" was a mistake based on the way people will react to such "failed" statements during the next event.

And I know I've seen it referenced before as that the "certain death" statement failed.

 

[rdale]

I see what you mean...

"Saying it failed, to me, would mean that a certain death statement was made and you still had thousands of people die. Maybe I'm just reading it wrong?"

No. Saying it failed, as all of the responses above have indicated, has more to do with the next time.

 

[Amymarie]

I agree completely that America is insane when it comes to lack of personal responsibility . I have no problem with a competent adult deciding to stay in a mandatory evacuation, accepting responsibility & the fact that no one is coming to save them. Just don't complain if it really is as bad as you were told.

Maybe you could post quotes from people who stayed in past hurricanes & it was as bad as they were told. The ones who lived have horrific tales & say they would never stay again.

Here is one thing that drives me mad-people who have small children & stay. I know it would be a legal & logistical nightmare to define & enforce, but I don't believe they should be allowed to keep their children in certain situations. It almost seems criminal for parents to keep children in areas forecasted to take a direct hit, at the least b/c of the almost surely dangerous conditions afterwards even in a miss.

For example, in regards to Ike, I heard an interview with a mother who said God was going to protect her & her children as they laid on the floor of their one story home, praying. Who would subject their children to this type of gamble?

 

[Jim Saueressig]

Well here is one posted right here in this very thread:
And I know I've seen it referenced before as that the "certain death" statement failed.

I used it based on the fact I had seen it somewhere a few times myself not that I think it failed, if anything it would possibly fail next time for those who refuse to leave the next event because they believe it was failed for Ike.

Personally I have no problem with the statement that was made I just think that it should have been worded different for the reasons I stated before.

 

[Scott Overpeck]

Thanks for some clarifications...

I completely understand the concern for the next time an Ike type storm hits the area. Personally, I hope it is a while before another strong Cat 2 or major hurricane hits the area. That said, I think now you have lots of people understanding what storm surge can do and its threat to life let alone property. There have been a few media interviews of survivors from Bolivar. All of them have said that the next time they will evacuate. I have personally heard from even people that rode out the storm inland that even they will get out as the 6-9 hrs of hurricane force winds was enough the scare the crap out of them. So, the hope is that in the future, IKe will be fresh in people's minds. They will be more likely to evacuate than before, and I hope the evacuation plans go smoothly as they did this last time. I would hope that the next time we would not have to use certain death type statements but as it is suggested, we could still use strongly worded call to actions like losing your house, etc. Hopefully there will be more people interested in preparedness and having a hurricane plan. We hope that we can educate the public more about surge, but sometimes it is just hard to imagine 10 feet of water in your house. People still have a hard time wrapping their brain around that. The hope is that they have a better grasp of that now and will be proactive enough to leave when evacuations are issued.

The other point that I'd like to make is the urgency of the situation. At the time of certain death wording, people were not evacuating. We knew that in less than 24 hrs areas of Bolivar and even the west end of Galveston would be inundated by surge and roads would be impassible. People needed to get out ASAP or they were not getting out. Some found that out the hard way. The problem that is arising is that we are trying to convey this message of urgency in a text forecast product that has to be issued ASAP. We have it set up that the HLS should not need much updating unless watches/warnings are posted which was the case at this time. We needed to get the warning information out so our canned call to action was what we had. We eventually toned it down on Friday when national media were latching onto it. I think what we are finding out is that perhaps the HLS is not the best product to disseminate such strong and urgent call to actions. I think perhaps this is where we could have partnered better with the media to get these points across that this storm surge presents a realistic threat on lives that live in surge prone areas. Certainly lessons learned for the next time, but if I had to throw out an excuse, we have not had to deal with a significant hurricane event for quite some time . The point being that perhaps we need to find better ways to get people to act given an urgency about a developing situation that threatens many lives.

Still, all that being said, if we go another 25 years without a major hurricane, we are right back where we left off. I hope we do not have to use certain death wording again, but if people are not acting, we may have to resort to it again. I do agree, the certain death type wording needs to not become common place and used in extreme situations. During Hurricane Ike, our office thought it was an extreme situation given the history.