man the problem with trying to explain to people why address level flood scores are so fraught is that i feel like i get backed into a corner where i have to start with like well what is knowledge really

Featuring a quote from @jdossgollin.bsky.social! As Adam notes, the core problem is these scores cannot be validated (in the strictest sense) for future use, because there are too many assumptions about future changes. We need to be transparent about these and look across a range of projections.

Tools that help savvy companies who hire scientists like @kellyhereid.bsky.social use, in conjunction with lots of other information, to answer specific questions are not intrinsically useful for ordinary people making the biggest financial decision of their lives

I am calling for a complete and total shutdown of daycare germs until somebody can figure out what the hell is going on (while I still have some backup care days left for 2026)

Climate risk science's greatest project successes have come in part due to transparency and reusability. We show that at the paper-level, there is plenty of room for growth. We talk about why this matters, what researchers can do, and how the greatest progress requires institutional support

This paper began as a manifesto but (largely thank to @crispapoll.bsky.social's hard work) evolved into an empirically grounded, thoughtful, and nuanced analysis of why transparency is *especially* critical for #climate #climaterisk research and how current standards fall short

Sorry editors -- I got about a billion invitations to review this week and had to decline most of them. Prioritizing (1) society journals, (2) open access journals, and (3) journals I have recently submitted to, down-weighted if I have more recently reviewed there.

All emails should cost $0.01. An email to all 20,000 people on campus should cost 200 bucks -- enough that you have to accept accountability and justify your choice, not so much that important stuff is blocked

Kelly's point is why I think the merits of Zillow removing climate risk scores (www.nytimes.com/2025/11/30/c...) is fairly nuanced. To make it clear: Zillow clearly did not do this for the right reasons, and the realty and developer lobbies have long been opponents of better climate information.

My favorite thing about Thanksgiving week is that all the companies who’s crap I need to unsubscribe from hopefully send me multiple reminders

Our free and open source software for estimating property level flood risk under uncertainty now has a peer reviewed software paper! We would love your thoughts for improvement and contributions! github.com/abpoll/unsafe @jdossgollin.bsky.social @vsrikrish.bsky.social

On Wed, Dec 10 I am offering a comprehensive introduction to Bayesian regression modeling, with exercises in both RStan and PyStan, at a steep discount in an effort to raise funds for World Central Kitchen and United Farm Workers. For registration details see betanalpha.github.io/courses/.

There seems to be very little coverage, but Hurricane Melissa is about to do terrible damage to millions of people. gdacs.org/report.aspx?...

Climate risk models require DEEP expertise to use well. It is easier every day to download some black-box risk estimate for your house/community/portfolio. However, the common mindset of "some data is better than no data" can lead to trouble!

Thanks to everyone who is out (on behalf of all of us knocked out by daycare germs) to demand and defend accountability, transparency, democracy, and the very concept of morality.

I literally only use this app to grumble

I knew having kids would change my life, but I wasn’t ready for Spotify to pack my recommendations with 98% white noise and nature sounds

📊New research published in #ERCL - in the 'Focus on #ClimateChange Informed Catastrophe Modeling to Support #ClimateRisk Management' 🌧️Bayesian spatiotemporal nonstationary model quantifies robust increases in daily extreme rainfall across the Western Gulf Coast: iopscience.iop.org/article/10.1...

Broke: I want AI to do my laundry and dishes so that I can do art and writing Woke: I want AI to fill out when2meet / doodle / calendly polls so that I can complain about bad papers on social media

Controversial take: every statistical analysis you do doesn’t need to be an R package. Packages should be infrastructure and set a very high bar of software engineering for them. To disseminate your analysis, just make some well documented templates and tutorials

What a clear, concise summary of the state and potential of renewable energy technologies. I particularly like how @rebeccasolnit.bsky.social includes the fact that nearly two-thirds of primary energy gets wasted today. Burning fossil fuels is a very inefficient way to provide useful services.

How exciting is research according to...

Grumpy take: you cannot just write in your methods section "we used X software", you should cite and note the software but you still need to explain what the software is doing and why it's a good choice

Talk about a profoundly shredded moral fabric

This country won't get serious about climate change until we're honest about flood and fire risk and insurance to match. Imagine if Camp Mystic had been required to carry liability insurance in the event that its campers were flood victims. Those cabins would have been moved years ago.

I understand everyone wants to cover the tragedy, but I would MUCH PREFER to be part of a public discussion of flood warning systems and safety BEFORE the next horrific disaster happens than AFTER. As long we remain uncomfortable with proactive safety measures, people will keep dying.

Early warning systems work. Early warning systems are crucial. This camp lost no one. See next skeet for quote. apnews.com/article/texa...

But this does illustrate a few tragic and uncomfortable truths. The first is that even quite good weather forecasts do not automatically translate into life-saving predictions--there's a lot of other work that has to take place to contextualize the forecast and ensure it gets to right people.

#Flood modelers: lots of chatter RE soil profiles in "flash flood alley". Do we really think floods would have been meaningfully different if you dropped that much rain on those hills, but they had more/deeper topsoils? Of course the effect is not zero, but not convinced it's big.

Based on some of the reactions and comments I am absorbing by average people and elected officials, I think there needs to be a serious explanation and lesson on what the limits of predictability are for floods like the one in Texas.

And don’t remove observational capacities and cut funding to weather forecasting agencies and then blame them for “faulty” forecasts!

Horrific things are happening along the Guadalupe River right now. Absolutely devastating for the families of everyone missing. I'm skeptical that we're going to build our way out of experiencing major flooding with 15+" rain on hilly terrain. Instead, we need...

I feel like I'm losing my mind. Here is the transcript of what Zohran Mamdani actually said about the phrase "globalize the intifada." This is his "refusal to disavow it." This is what has people terrified. Just fucking read it. Source: podscripts.co/podcasts/the...

My latest feature is out: As the world has warmed, the regional flaws of climate models have grown apparent. Meanwhile, methods to localize the results of models have proliferated. It all leads to lots of uncertainty -- within a very certain global trend. And cities need answers now.

🔴⚠️⛈🇦🇺Unprecedented rains, almost 500mm in three days for some towns, have hit #NewSouthWales with floods that involved 50000 people. #Australia declared the natural disaster for the state.⬇️#Sentinel1 animation of the flooded area covering 160km2.

We are living in one of the few times in human history where you can buy a loaf of bread and if it says it’s made with wheat flour and salt and yeast and water and nothing else, it’s probably made from just those ingredients, and some people want to END this.

“Hey Meta, give me the shallow illusion of a personality” is my favorite ad campaign right now

NSF cancelled our flood justice network today. I’m of course devastated. But the work will continue. We were already creating a network and we won’t stop now! @socialpixel.bsky.social P.S. there’s nothing woke about creating solutions for communities that flood. And saves the govt money!!!!

QEPD

Lots of other things going on and geopolitics is way beyond my expertise but glad to see some mainstream folks pushing back on the disturbingly widespread desire for another Cold War overcast.fm/+ABLShO7PaNM

The plan “…eliminates NOAA's Oceanic and Atmospheric Research (OAR) Office and its 10 research laboratories and 16 affiliated Cooperative Institutes…one only needs to see what NOAA research has provided to the U.S. taxpayer and imagine where we would be without…” www.ametsoc.org/ams/about-am...

Cancelling an excellent scientific program because their findings are displeasing is NOT a good sign

An extraordinary separate dissent from Justice Jackson accuses the majority of trying to pass off its dirty work under the cover of the shadow docket and explicitly invokes Korematsu, the Japanese internment case. Wow. www.documentcloud.org/documents/25...

If I have no representation do I still have to file taxes?

New JP Morgan Climate Risk report out and even they are calling the insurance/property crisis the "doom loop" now

The story of this man's abduction and deportation to a brutal El Salvador prison, without any due process—and apparently by mistake!—is the stuff of nightmares, literally ripped from some dystopian fiction. Kafka or Orwell. Or Stalin's Russia. Hitler's Germany. Or apparently, Donald Trump's America.

The frustration when you agree to review a paper because the abstract seems relevant and exciting, but then you download it and it's more 😬

Yes, I am aware that AI is often incredibly inaccurate, and that these powerful assholes knowingly use it as an unaccountable and inscrutable force to validate whatever they wanted to do anyway, that is *literally the point I’m making*

If you are visiting the United States, having a negative view of Trump is enough to get you denied entry. If this was coming out of Russia or China, we would deplore such an authoritarian attack on freedom of expression.