Practical Engineering
2 HN recommenders · recommended in 5 months, 2017–2025
7:45 · 2.3M views · Mar 7, 2016
Practical Engineering explains what a 100-year flood actually means, combining statistics, hydrology and how society treats risk. Grady walks through local drainage infrastructure and the caveats in estimating flood probability.
HN points to it as a clear explanation of what flood return periods really mean.
5 comments from Hacker News
Even that is a little misleading (given how bad our intuition about statistics is). A '100 year flood' is flood that has a 1% chance of occurring in any given year (it's independent of what happened 1, 5, or 50 years ago). Calling it a '100 year flood' makes people think 'oh, good, now we shouldn't have one of those for another ~100 years'. In reality, the odds of a flood of the same magnitude occurring next year are exactly the same as they were this year. Grady (of the Practical Engineering YouTube channel) has a great video on this topic: this video (Grady is in the Austin area, so hopefully he hasn't too much 'practical' experience from this storm...)
It's not human action that would worry me, it's nature. Freezing cycles could easily erode a pool in a few decades and make micro-cracks that fester unnoticed for years. Rust is an ever present issue for re-inforced concrete and is the cause of countless civil engineering disasters. Practical Engineering on Youtube is a great overview of the issues we humans have with water: [LINK] Of interest here are these videos: [LINK] [LINK] this video
Just for clarity, you're claiming that there's data to support that 10-year flood plain areas should be marked as having a 100% chance of having a flood every year and you have enough data to say a 100-year flood plain should be reclassified as a 10-year flood plain? I ask this because those numbers are shorthand for event probabilities given (flawed) observation data over large time-frames. Argue all you want about climate change moving these number due to rising sea levels. I agree it's happening. That being said, hyperbolic assertions such as this don't help. The reality is that our flood plain designations are very flawed, but they are the only current metric we have to assess some of these risks. Every location that receives precipitation or is near water has some risk of flooding and that risk of flooding may be hidden. The whole state of Florida is low elevation. Something like 1/3 of it is very low elevation and subject to sea-level rise causing catastrophic changes in coastline. But the other 2/3 of Florida is something like 100ft (30m) above sea level. The other issue with Florida is that it's the flattest state in the US. That flatness contributes to flooding being difficult to properly predict. > Dont buy houses in flood planes. I'd just like to call out this hard-line assertion as well. Have you looked at a flood plain map? They have hard borders of zones. Those zones are calling out that event probability. If I buy a house 10 ft outside that zone, do you think m
> I observe no denigration in his comment. These people made a choice. It's not even my personal opinion. I am merely describing the biases of the general public that result in the discrepancy in newsworthiness. It's not due to some coastal conspiracy against inland states. Florida is viewed by many as a backwards state, a popular target for mockery and derision, so a bias against inland states does not explain why hurricanes in Florida get more news coverage than floods in inland states. To reiterate myself a third time, the discrepancy arises from a perceived , not actual difference in culpability of the victims. People who are hit by hurricanes are thought to be the victims of bad chance, while the people living on floodplains are thought to be people who knowingly tempted fate. That this is not squared up with reality is of little consequence because newsworthiness is not some objective trait; newsworthiness is the product of the subjective biases and worldviews of the general public. I'll offer another reason for why the discrepancy exists, again relating to hurricanes being named. When a community gets hit by some hurricane, they're not hit by some generic hurricane but rather by a particular hurricane, that had a name and it's own news cycle leading up to contact with the coast. And throughout that news coverage, the location that the hurricane would come to shore at was up in the air. The weather forecasters were offering up stochastic predictions, but none knew for c
"Relatively easy" is brushing over an awful lot. [LINK] You can get a rough approximation, and that's been done at great effort for the entire US at the FEMA website: [LINK] . It's not a simple problem to solve. Some of the caveats are discussed in this Practical Engineering video: this video Volume of outflow through a stream is equal to rainfall times watershed area. Seems intuitive, right? Except what about water that gets absorbed into the ground, or captured in a temporary holding pond, slowed by vegetation, or accelerated in a storm drain...I think Grady put it well when he pointed out that the silly Imperial units of outflow in cubic feet per second, rainfall in inches per hour, and watershed area in acres multiply together to 1.008, which is off by about 1% but close enough that civil engineers just throw away the constant entirely because it's such an imprecise approximation that the error is irrelevant.