MikeB4: Sorry? The thread is about Tsunami alerts? You have lost me?

No, it's about the fact they are issued all the time with no backing from anyone except someone with access to a social media/twitter account.

Read the first post again, and maybe re-read it.

MikeB4:

When I owned boats I experienced it at Napier and South Wairarapa

Neither of which were in Auckland, last time I checked.  And so don't benefit from the sheltering effects of the Coromandel Peninsula, GBI and LBI - which all make a hazardous Tsunami in Auckland very unlikely.  For example, I left home for work on 13 Nov and there was an extant Tsunami warning for Auckland at the time, but I still drove to work on the waterfront, as did all my colleagues.

The CD seem to totally ignore the geography/oceanography of the Auckland region when issuing their alerts, which degrades their value

Next time there's an evacuation alert for my area (South New Brighton, Chch - not much elevation), OR the sirens go off (The two astonishingly are not linked!) I'm not going to immediately comply. I'm going to go online and figure out for myself if I think there's a risk. If it's low tide then pretty much nothing could get me to move... If it's high tide when the wave is predicted to hit then I'll think more carefully...

Why not listen to the experts you say? Well, whoever is in charge of the sirens in Chch couldn't organise a piss-up in a brewery, and in every instance I can recall in Chch the alarms/alerts have proven so uselessly hyperbolic that my current working assumption is one of the following.

1) They are following a badly flawed model

2) Their model has terrible input data

3) They've consciously decided to report some comical number of standard deviations on the "OHMYGODTHEWORLDISENDING" scale - meaning that if their model (say) predicts a 50% chance of a 20cm tsunami, a 5% chance of a 1m tsunami, a 1% chance of a 2m tsunami and a 0.1% chance of a 3+metre tsunami, they are going to pick the 3m tsunami to use in the warning.

I can't think of any other possibilities. Of course I'm not as qualified as the people doing this for their job - and I'm certainly not as smart as some of them either, but when they consistently serve up such a chasm between predictions and observed outcomes, you don't need to be an expert to figure out that they are doing something terribly wrong.

Now perhaps what they are doing wrong is clearly communicating to the public what their warnings mean - and I'm sure there are some colourful discussions in the various departments Responsible For This Sort Of Thing - but how would people react if they said they do in fact use the absolute worst case model (say 1 in a thousand outcome) to issue the alerts and warnings?

Me? I love data. Serve me up bad data and I'll stop paying attention to it.

Cheers - N

Talkiet:

 

Next time there's an evacuation alert for my area (South New Brighton, Chch - not much elevation), OR the sirens go off (The two astonishingly are not linked!)

 

 

 

One of my main issues!!!

Talkiet:

 

I'm not going to immediately comply. I'm going to go online and figure out for myself if I think there's a risk. If it's low tide then pretty much nothing could get me to move... If it's high tide when the wave is predicted to hit then I'll think more carefully...

 

Why not listen to the experts you say? Well, whoever is in charge of the sirens in Chch couldn't organise a piss-up in a brewery, and in every instance I can recall in Chch the alarms/alerts have proven so uselessly hyperbolic that my current working assumption is one of the following.

 

1) They are following a badly flawed model

 

2) Their model has terrible input data

 

3) They've consciously decided to report some comical number of standard deviations on the "OHMYGODTHEWORLDISENDING" scale - meaning that if their model (say) predicts a 50% chance of a 20cm tsunami, a 5% chance of a 1m tsunami, a 1% chance of a 2m tsunami and a 0.1% chance of a 3+metre tsunami, they are going to pick the 3m tsunami to use in the warning.

 

 

 

I can't think of any other possibilities. Of course I'm not as qualified as the people doing this for their job - and I'm certainly not as smart as some of them either, but when they consistently serve up such a chasm between predictions and observed outcomes, you don't need to be an expert to figure out that they are doing something terribly wrong.

 

Now perhaps what they are doing wrong is clearly communicating to the public what their warnings mean - and I'm sure there are some colourful discussions in the various departments Responsible For This Sort Of Thing - but how would people react if they said they do in fact use the absolute worst case model (say 1 in a thousand outcome) to issue the alerts and warnings?

 

Me? I love data. Serve me up bad data and I'll stop paying attention to it.

 

Cheers - N

 

 

 

And the rest is exactly as I feel, but I have no say in things with my kids living there and me being in Oz.

The fact there was an evacuation order, but they didn't bother to tell anyone that wasn't on social media at the time (difficult while calming kids, no power, overloaded network, etc) shows they have NFI what they're doing. And then rescind the order hours later, again only on SM.. 

It's a cluster eff

blakamin:

 

 

 

But anything over a 6 in the pacific region pretty much "automatically" gets a tsunami warning from civil defence... WHY?

 

Because it's new science. Remember that before the Indian Ocean tsunami, there was *no* warning system anywhere. People hardly even knew the word.

So they've built a new system which generates warnings where there is some threshold probability of an actual tsunami. Someone has fairly arbitrarily decided what level of risk should generate a warning. And, naturally, they've erred on the side of caution. As time goes by and the risk factors are better understood, and more and better instrumentation is put into place, the quality of the data will improve, and so will the reliability of warnings.

frankv:

 

blakamin:

 

 

 

But anything over a 6 in the pacific region pretty much "automatically" gets a tsunami warning from civil defence... WHY?

 

 

Because it's new science. Remember that before the Indian Ocean tsunami, there was *no* warning system anywhere. People hardly even knew the word.

 

So they've built a new system which generates warnings where there is some threshold probability of an actual tsunami. Someone has fairly arbitrarily decided what level of risk should generate a warning. And, naturally, they've erred on the side of caution. As time goes by and the risk factors are better understood, and more and better instrumentation is put into place, the quality of the data will improve, and so will the reliability of warnings.

 

 

 

How does Australia and the US seem to get it right? They both say straight up "No tsunami threat" but NZ issues a warning, for the whole country, and a full 20 minutes later... 

blakamin:

 

Is it just me or are there tsunami warnings being issued every time there's an earthquake in the wider pacific region?

 

This morning there was an quake near Papua NG and, as usual, NZ issues a tsunami warning.

 

 NZ had a warning in place for at least an hour, and it took 20 minutes (maybe) to issue a warning, after Australia said there was no risk (I'm still yet to see an actual warning in Oz).

 

Is this going to end up with a case "The Boy Who Cried Wolf"?

 

Are they just issuing warnings after every quake over 6?

 

Thoughts?

 

Only asking as my kids are about 400m from a beach and in a tsunami area (even though there was a warning and evacuation issued for their area after Kaikoura, they didn't even know).

 

They are right to issue a warning after a big quake. They may not know what exactly might happen......so they are telling people to be careful and stay away from the water's edge...and be ready to head for higher ground.

It's prudent. If they wait until they know for sure, any chance of actually warning people might be gone.

It IS a warning....be careful. 

Quakes can cause odd, strong and unpredictable currents around coasts....even if the sea doesn't inundate the land. 

blakamin:

 

 

 

How does Australia and the US seem to get it right? They both say straight up "No tsunami threat" but NZ issues a warning, for the whole country, and a full 20 minutes later...  

 

OZ is over 200km west of us. It may well be that no wave is headed that way.....while at the same time that might not be true for us. 

Maybe the ocean buoys detecting the waves headed our way are 20 minutes further away in our direction versus where they are for OZ.....from a given quake event.

It's going to be complicated and variable. :-)  

blakamin:

 

frankv:

 

blakamin:

 

 

 

But anything over a 6 in the pacific region pretty much "automatically" gets a tsunami warning from civil defence... WHY?

 

 

Because it's new science. Remember that before the Indian Ocean tsunami, there was *no* warning system anywhere. People hardly even knew the word.

 

So they've built a new system which generates warnings where there is some threshold probability of an actual tsunami. Someone has fairly arbitrarily decided what level of risk should generate a warning. And, naturally, they've erred on the side of caution. As time goes by and the risk factors are better understood, and more and better instrumentation is put into place, the quality of the data will improve, and so will the reliability of warnings.

 

 

 

 

How does Australia and the US seem to get it right? They both say straight up "No tsunami threat" but NZ issues a warning, for the whole country, and a full 20 minutes later... 

 

 

 

So immediately after a quake offshore, its known its a slip EQ, therefore no risk or an uplift, which has risk?? No. Not anywhere near close. . How do you wish to manage the risk? Most of the time its ok so lets most of the time go with the stats??  Or, take the safe approach, IMMEDIATELY issue a warning, as it might be a risk. When its not a risk, remove the warning. Or, lets wait an hour?? Should an uplift  EQ occur in or near the Kermadec trench its too late if you spend time evaluating the risk. 

Linuxluver:

 

blakamin:

 

Is it just me or are there tsunami warnings being issued every time there's an earthquake in the wider pacific region?

 

This morning there was an quake near Papua NG and, as usual, NZ issues a tsunami warning.

 

 NZ had a warning in place for at least an hour, and it took 20 minutes (maybe) to issue a warning, after Australia said there was no risk (I'm still yet to see an actual warning in Oz).

 

Is this going to end up with a case "The Boy Who Cried Wolf"?

 

Are they just issuing warnings after every quake over 6?

 

Thoughts?

 

Only asking as my kids are about 400m from a beach and in a tsunami area (even though there was a warning and evacuation issued for their area after Kaikoura, they didn't even know).

 

 

They are right to issue a warning after a big quake. They may not know what exactly might happen......so they are telling people to be careful and stay away from the water's edge...and be ready to head for higher ground.

It's prudent. If they wait until they know for sure, any chance of actually warning people might be gone.

 

It IS a warning....be careful. 

Quakes can cause odd, strong and unpredictable currents around coasts....even if the sea doesn't inundate the land. 

 

 

 

Exactly. It a warning, not a statement that a tsunami will arrive at 1034 hours

Not sure which traffic jam I should join next earthquake.

Last one the traffic was snarled in kapiti at midnight with no warnings.....probably better to just buy a long snorkel

blakamin:

 

frankv:

 

blakamin:

 

 

 

But anything over a 6 in the pacific region pretty much "automatically" gets a tsunami warning from civil defence... WHY?

 

 

Because it's new science. Remember that before the Indian Ocean tsunami, there was *no* warning system anywhere. People hardly even knew the word.

 

So they've built a new system which generates warnings where there is some threshold probability of an actual tsunami. Someone has fairly arbitrarily decided what level of risk should generate a warning. And, naturally, they've erred on the side of caution. As time goes by and the risk factors are better understood, and more and better instrumentation is put into place, the quality of the data will improve, and so will the reliability of warnings.

 

 

 

 

How does Australia and the US seem to get it right? They both say straight up "No tsunami threat" but NZ issues a warning, for the whole country, and a full 20 minutes later... 

 

 

 

Pumpedd:

 

Not sure which traffic jam I should join next earthquake.

 

Last one the traffic was snarled in kapiti at midnight with no warnings.....probably better to just buy a long snorkel

 

Quite a few in Kapiti who tried to flee couldnt get their garage door to stay open with no electricity.