Dave Zatz

March 2026. Midjourney v8 still can't do hands 😅

⚠️ Why the storm forecast that scared D.C. fell short Monday brought one of the most ominous severe weather forecasts in years, including a rare Level 4 out of 5 risk for damaging winds and tornadoes. Schools dismissed early and plans changed but while there were indeed many reports of severe weather (111 severe wind reports), some of the most feared impacts never materialized. Here’s what happened: • Limited storm fuel: Thick cloud cover held temperatures down, reducing the instability needed for widespread severe storms • Early storms “used up” energy: A late-morning line likely “stole the second round’s thunder” • Downdrafts couldn’t get going" Moist air near the surface limited the ability for strong winds aloft to reach the ground At the same time, the concern was justified. The setup had the ingredients for dangerous storms, and had those ingredients come together differently, the outcome could have been far worse. But this event also highlights something important: 👉 We need to do a better job communicating uncertainty. Severe storms are highly sensitive to small changes. Even with strong signals, they’re not guaranteed. That nuance didn’t come through clearly enough. Forecasting severe weather remains one of meteorology’s hardest challenges — and this was a reminder that how we communicate risk is just as important as the forecast itself.

Ce matin, je suis à Disneyland Paris pour l’avant-première de l’extension Frozen. On a eu 10 minutes avec le robot Olaf et un des ingénieurs qui l’a conçu. Bientôt un reportage pour expliquer la technologie sur Numerama !

I also may be a poor role model.