🚨🔥 NEW: Increasing mortality from wildfire smoke is directly traceable to greenhouse gas emissions. In our new paper in @pnas.org , we show that every megaton of CO2 emitted costs the US more than 10 million dollars in health costs from wildfire smoke. www.pnas.org/doi/10.1073/...
Posts by Iván Higuera-Mendieta
📈🚨NEW: Extreme heat events are increasingly threatening to become mass mortality events. In @natclimate.nature.com today, we project tens of thousands of deaths in a single week across Europe if extreme heat domes coincide with rising global temperatures.
www.nature.com/articles/s41...
We are excited to announce the release of the Environmental Hazard Adaptation Atlas, an effort to map ongoing and future environmental hazards and their impacts on society, and to provide up to date evidence on what policies and interventions work to reduce impacts: adaptationatlas.org. Quick thread
This is well researched. But I bristle at the idea of this as a “scientific standoff.” A hallmark of science is that there are almost always contrary paradigms challenging the dominant one. But you have to know when the weight of evidence is on different planets.
news.mongabay.com/2025/08/old-...
A large low over North America guides smoke over the skies of the eastern US, originating from wildfires on the western side of the continent.
Does prescribed burning reduce overall smoke from wildfire? We have 2 new papers that try to quantify. Answer: each acre Rx burned yields ≥3x more reduction in future wildfire smoke than is emitted in the Rx burn. But can take years to realize benefits.
www.stanfordecholab.com/blog/prescri...
"just immigrate legally bro"
Ghost of christmas past can help you test whether you need to worry about negative weighting in your event studies
nbviewer.org/github/apoor...
Coming soon to a pyfixest and preprint server near you
github.com/apoorvalal/T...
I actually think the most dangerous thing with AI is vc/McKinsey types overestimating the abilities of AI and using it in places where it shouldn't be used. Dumb and dangerous
Be the you you want you to be.
AI in weather forecasting?
Check out this amazing thread diving into the robustness of modern data-driven weather forecasts in past and future climates.
(Honestly, one of the biggest questions about ML in weather in in- and out-of-distribution events.)
🧪 #mlsky
Samples y | x from Treeffuser vs. true densities, for multiple values of x under three different scenarios. Treeffuser captures arbitrarily complex conditional distributions that vary with x.
I am very excited to share our new Neurips 2024 paper + package, Treeffuser! 🌳 We combine gradient-boosted trees with diffusion models for fast, flexible probabilistic predictions and well-calibrated uncertainty.
paper: arxiv.org/abs/2406.07658
repo: github.com/blei-lab/tre...
🧵(1/8)
"Love is for the ones who love the work" - excuse me while I take a moment to catch my breath
I would totally subscribe to a premium tier that includes HTML & CSS profile customization à la MySpace circa 2005.
This is so helpful to navigate across the n starter packs around here.
btw this is why friends dont let friends skip the “boring classical ML” chapters in Elements of Statistical Learning‼️
(True story: the origin of this case study is that @alanjeffares.bsky.social[big EoSL nerd] looked at the neural net eq&said “kinda looks like GBTs in EoSL Ch10”&we went from there)
after years of opposing the owning-a-house part of the American dream, i've finally been convinced otherwise.
Before I used scikit-learn, TensorFlow and etc. today I only use Requests…
@daisylu.bsky.social 💘