Hacker News Reader

Comment by PlunderBunny | Hacker News Reader

I wonder how frequently that river (and the rest of the world) will experience once-in-a-hundred-year weather events from now on.

View on HN

The "Once in a hundred year" saying is misunderstood. It's actually 1% chance each year. So you roll your D100 every spring.

Whether that's an acceptable risk is up to you. Having lived in central Texas, it's a region known for it's flash floods, and you should take the warnings seriously. If there's heavy rainfall - you should be asking where all that water will go.

In this case, I hope that Texas looks into upgrading their alerting system so that people get word in time to leave the area. The news was saying the weather center added additional staff for the storm (five meteorologists) but if their forecasts can't get to where they're needed, that's not good enough.

> In this case, I hope that Texas looks into upgrading their alerting system

Since the July 4 flooding, I feel like the over correction of flood warnings via mobile phones is not a good thing. My phone has been going off with near daily flood warnings with 0% precipitation. I get how rain upstream can cause flooding where it’s not raining downstream, but it hasn’t even been raining upstream. I’m also well over 200 miles away from where the worst of the flooding happened. People that far away do not need these warnings. The notifications have listed counties not close, so it just comes across as “let’s do something just to say we did something”

Mobile notifications is a terrible solution to this

Need ground alert sirens or something that would be taken seriously

It's a terrible solution if it is the only solution. Up until this situation, the alerts I received on my phone have been pretty spot on, which has been impressive. Yes, local sirens are a good idea, but they come with caveats. I have local sirens in my area, but they are for tornado or other severe weather. If I hear those sirens, my action to take is totally different than for a flood. Naturally, if I were to come to an area where the sirens are meant for flooding, my reaction to them would be the wrong move. I would hope that a siren for flooding would just sound different than how the established tornado siren warning system sounds.

This trained response to a siren/warning system is the reason they chose not to use the tsunami warning system in Hawaii during the fire. When that siren goes off, people seek higher ground which would have driven them to the fire.

This might be a good project for someone - take the terrain contours and rainfall location & amount into account when determining where to send alerts.

I think once-in-one-hundred-year is a definition for weather events used in building standards (sorry to be vague and possibly misleading - if my resident architect/partner was here, I could be more specific). I've certainly heard of constructions like levies etc "build to withstand once-in-a-hundred-year storms". I wonder if these standards are being revised by appropriate international bodies?

I think it has become a "term of art" used by agencies like FEMA and insurance companies. And it makes for a good soundbite on the news.

But you would question the dice if you roll your D100 ten times and it shows the same number on six throws

At a (central Texas) city council hearing today on granting someone a variance to build tennis courts on a meadow next to a creek, a longtime resident said, “I think I’ve lived through five 500-year floods here.”

Where I live in Texas, we had three “500 year” events in three consecutive years, with the last one probably closer to a 1000 or 10000 year event - more rainfall in one spot than anywhere else ever measured.

We’ll see how long that record holds (I’ll take the under on 1000 years).

There's probably many other 1 in a 1,000 year events that haven't happened in many thousands. It's really hard to reason about infrequent events, especially post-hoc.

If in the Netherlands a river floods millions are affected and there is trillions in damage. If in America a river floods a few hundred people die. That's why my country takes this stuff seriously.

I guess if you’re building something in vulnerable spots tennis court is better than house

Not for the neighbors. The courts are 10x the impermeable cover of the house they’re also trying to build. There should be nothing on that land.

There are over 2000 watersheds in the US. It would be unusual if we didn’t see around 20 100-year floods every year.

I would think they'd be correlated enough to mess up the numbers a little.

Right. It's the multiple-endpoints principle. The extreme events feel overly frequent because we cherry-pick and only notice them. We never notice the other 99/100 of places that don't have a hundred-year flood in a year.

Extreme precipitation events are also getting more common.

Reply on news.ycombinator.com