But main message that isn't getting through is that if you have to choose your building level mitigations - this says zone zero first, hands down. That's big news.

Feels like we've hit "bargaining" stage.

All around bad news for foundation garden lovers

Oh noo I've heard that too, "well if plants were just watered they won't burn!" 1, wildfires more common in drought when probably under watering restrictions anyway, but 2, just no.

Have you gotten tagged on the "native plants are ember catchers that will stop embers from reaching house!" one yet?

Bet it's going to make you *real* popular in Berkeley though... just starting on zone zero enforcement in high risk areas, but people really love their gardens here.

Yeah my jaw hit the floor when I saw difference in efficacy. If it holds it really makes me question the heavy building focus in CA - sure it helps a little, but putting gravel or a patio around your house is so much faster to deploy at scale.

Cc @michaelwara.bsky.social you'll be interested too.

Ahhh yeah good call on duration, would be fun to link with RI driving those longer lived high cat storms too.

As you say, lots of ways to dig in.

I'd imagine getting a good sample size with high enough confidence in pre fire state probably a challenge for modeling, but it sure is intriguing enough to want to dive in more.

@gollnerfire.bsky.social am I missing a reason why y'all aren't shouting this result from the rooftops? Says to me that much cheaper veg mitigation is *way* more effective than expensive project of whole building retrofit.

Cc @cfarivar.bsky.social - also WAY drives up value of inspection services

Oh dang this is cool - in a sense really shows how unusual the drought was.

Have you tried splitting this up by genesis location at all? Not to go all "blame ENSO for everything" but La Niñas dropping storms in the CB may stick some interesting decadal variability in here, upping landfall prob...

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....

Aw dang, I hadn't seen this paper, this is really cool!

Also a *really* strong argument for making sure gages stay in place to collect that flood of record...

Nothing immediately obvious comes to mind, although could visualize something like a long historical river gage where you run an extreme value fit to 100yr RP as you gradually add each annual peak and see how it varies as the record lengthens...