Good for them. And great for you - congratulations Madison!

When I served on the geosciences DG panel, my all-time favorite comment was from a very senior prof who said that he chose NOT to publish in science or nature because their article formats were too short to accommodate quality science.

Me too, I said to the other panelists, me too ...

It was and remains one of my most unloved papers. But I had a good time working on it with Manfred. We had a paleoflood version cooking in the back burner for a good while (“You call that a flood? That’s not a flood”), but alas it probably will never be.

Yeah, I think the problem is right for some interesting simulation exercises. A colleague and I did something a little bit similar a few years ago where we knocked out the peak flow from every USGS gauge in the country and then re-computed the FFA: onlinelibrary.wiley.com/doi/epdf/10....

There should be a better source that explains why low probability events have so much wiggle room tied to them. @kellyhereid.bsky.social do you have a go-to source to explain why you should never turn your back on the 100 year flood?

In this example from Bulletin 17C, the estimated discharge for the 100-yr flood (aka 1% AEP) is somewhere between *squinting* 4,000 cfs and 20,000 cfs. Even with an actual observed flood close to 10,000 cfs to act as a reality check for the estimate, that is just a massive range.

In the US, the go-to reference for flood frequency analysis is Bulletin 17C from the US Geological Survey: pubs.usgs.gov/tm/04/b05/tm...
My point about the uncertainty in the 1:100 year of event is not stated explicitly, but it’s obvious if you know what to look for.