Hacker News Reader

Comment by crote | Hacker News Reader

> While Venture Global declined to make officials available for interviews, it has insisted its steel wall is designed to withstand a once-in-500 years storm.

The problem with those "once-in-500 years" figures is that they are based on historical data - and climate change is rapidly invalidating that data. Climate change doesn't mean it's 1.5°C warmer year-round, or the sea level is 20cm higher worldwide: it means weather becomes more more extreme. What was a "once-in-500" event a few decades ago might turn into a "once-in-25" event a few years from now. We are already noticing those changes in day-to-day life!

View on HN

It’s not really a problem, you just adjust the size of your assumed storm. We have lots of climate models and data to adjust predicted sizes.

Climate change effects are already being included when calculating wind and wave loading in many codes.

The real issue is that engineering codes use frequentist methods which make it hard to consider uncertainty, which often makes it unclear what the real safety factors are. This issue is being solved by using probabilistic engineering techniques, and in future, more sophisticated causal inference.

Climate. Models aren't really specific enough to predict a new 500 year storm in a specific location.

Those thresholds and definitions are based on the data record, and already encoded into regulation and a 100 years of construction.

What we see instead is Regulators simply increasing the requirements from a X year storm to a 2X year storm, and leaving the definitions. This is what I have seen with the California building code

You're on the right track, but your framing is still off wrt how the engineering design process works.

Assuming that designing for a 500-yr storm has anything to do with 'predicting what a 'future 500-yr storm' (or 25-yr or 100-yr) looks like is dead wrong. Irrelevant.

The 'definitions' are not left alone, they are updated as time goes on. But with historical data, and they are not extrapolated/predicted out into the future.

Engineers (PEs) design by taking known criteria and then applying probabilities and factors. They do not predict criteria. It's a subtle but important distinction.

A 500yr event, by definition, is actually the one year probability of a 1/500 chance event.

And it's up to the designing engineer to choose and state whatever the assumptions are that go into that.

But a levee designed this year will use this years current 'storm definition' just as it uses this year's building code. Not a future one.

(Sometimes the storm/ event definitions seem stale because things like flood maps might only get updated every few decades.)

> It’s not really a problem, you just adjust the size of your assumed storm.

Adjusting how you call the storm doesn't make the wall bigger: that's the problem. (It also makes their statement untrue in the present day, regardless of if it was true when the wall was designed.)

It's not 'adjusting how you call the storm'.

It's the engineer defining his design criteria (the 'design-storm'), based-on and benchmarked-to local historical data, including recurrent intervals.

The wall doesn't need to be bigger if next years data changes. It was designed for a (this year) 100yr- or 500yr-storm, not a guess of hypothetical future one.

The real problem I have with the "X-hundred year Y" things is that it leads even technical and mathematically inclined people to get confused, let alone people who are neither. It makes it sound like it happens on a schedule, or that it is not common / a minuscule risk.

A "100 year flood" implies there's a 1-.99^30 or a 26% chance of happening over the life of a mortgage.

I always feel like if it were phrased in that way people would grasp the concept better. Same idea holds for any X and Y.

These make for great entertaining/ filtering threads though.

Obvious who knows probability and who's a PE, plus various levels of self-awareness. Otherwise intelligent ppl taking terms at face value that are actually terms-of-art / industry lingo. And sometimes grossly misapplying such terms.

And that's before you even get into politics/ ideology, should you want to judge such things.

It's a 6d personality/intelligence grid.

Throwing 100 interviewees into a hypothetical thread like this is the closest I can think of to the perfect early round hiring process.

And considering weather in the US has been recorded for only about 250 years it’s hard to say what a 500-year storm even looks like.

A "100 year storm" is a storm with a 1% probability of occurrence within any given year.[1] So a 500-year storm has an 0.2% annual occurrence probability.

Those of us who understand probabilities immediately recognise that even assuming the base-line probabilities are correct, this will occur far more frequently than once in 100 years.[2]

And of course, the probabilities require a known historical distribution, which itself changes over time (climate change being only one of multiple factors). It is possible to make inferences based on the observed rate vs. intensity or magnitude of events. In many cases these follow an inverse-log relation, such that order-of-magnitude greater intensity events occur with an order of magnitude less frequency. Tracking the more-frequent, lower-intensity incidents actually gives a good proxy for how often higher-intensity events will occur.[3]

________________________________

Notes:

1. "The 100-Year Flood" <https://www.usgs.gov/special-topics/water-science-school/sci...>

2. A 1% probability event has a 99% probability of not occurring. For a period of time, raise 0.99 to the power of the periods occuring. Over a 100 year period, that's 0.99^100, or a 63.4% chance of no such events occurring. For the probability of k occurrences over n periods you'd apply the binomial distribution: <https://en.wikipedia.org/wiki/Binomial_distribution>.

3. This is a relationship which applies to a whole host of events, and is very frequently observed in general physical phenomena. Asteroid impacts, earthquake intensities, volcanic eruptions, storms, wildfires, terrorist attacks, and the like.

It's plenty, because 30yrs is usually the rule of thumb for enough, and extreme events are more likely to be recorded even in the absence of/before there were regular record intervals.

There are well established probability distributions for these things, so you're really just solving for which published curve and maybe a magnitude multiplier and seeing where your data fits.

Also there's uncertainty baked-in and accounted for at every step of the engineering design process.

Eg. Geotechnical engineers love to calculate 6 digits to 4 decimals then throw in a 400% factor of safety.

>it’s hard to say what a 500-year storm even looks like.

Not to worry, that was taken into account when they came up with the 500-year naming scheme.

Sure it sounds like 500 consecutive years, but it's really a 1 in 500 chance of absolute devastation every year.

How bad was the devastation?

Approximately 5x as bad as the ones where it's only a 1 in 100 chance of striking every year, naturally.

This is like saying that if you've only seen 250 people in your life, it's hard to imagine someone who is at the 99.8th percentile of height.

It is though? You probably wouldn't be aware that hormonal disorders can exist and affect height, so you'd massively underestimate how tall people can actually get.

Now I want to know how those people fit in a distribution. It could still be a continuous function.

Of course it's continuous, but (guessing) it's closer to power-law than normal - the tails are fatter than you'd think, and when you try to estimate variance from a sample you get an underestimate.

You are misunderstanding the engineering language in a term-of-art way.

1 in 500 year storm is a one-year probability of 1/500, based on now.

There is no implied or expected future change that needs to be considered. That's already baked into selecting the design-storm-interval of 1 in 500 vs 1/250 or whatever else.

A 1 in 1000 year event need not be, and likely won't be, the same/valid probability in the year 3024. Or even 2074. And that's OK. The engineers know that.

How is the one-year probability calculated?

Historical data only.

Rule of thumb is usually 30 yrs minimum of records/data from usgs/ nws or similar weather stations. But it's often less.

Then it's just a probability eqn, and there are various rules for making adjustments based on things like distance away from the weather gages.

But the more fundamental thing is you are setting a static design benchmark, not a future or perpetual target.

Once that benchmark is set, you apply factors/probabilities of various failure modes.

Or you can just choose a larger design-storm.

But you don't say 'I'm designing for what the 2025 or 2074 or 2124 100-yr storm will be'.

> 1 in 500 year storm is a one-year probability of 1/500, based on now.

It's more likely to be based on historical data that as of now is outdated and unreliable

It is based on only historical data, by definition.

It cannot be 'outdated and unreliable' (due to future expectations), because that's not how it's defined.

And it's not how engineering design processes work. (Google LRFD or ASD)

'Now' means the historical data that is current as of design.

You are merely setting a design criteria benchmark. Based on historical data.

If you want to account for future uncertainty, it goes into the other side of the design equations, via the factors and probabilities.

But more fundamentally, picking a 1/500 design storm already takes into account certain aspects of risk levels, and has nothing to do with expected lifetime or survivability. You can and do often pick design storm intervals that greatly exceed design-lifetime, for various reasons.

As a marketing slogan, withstanding "once-in-500 years" storms is intended to (a) disconnect the listener's brain from the reality of climate change, and (b) ensure that in a decade or so there will be repeat business when it has to be replaced/upgraded.

Reply on news.ycombinator.com