Plus, the AI hurricane model that’s proven its worth
Saharan air dampens the early hurricane season, an AI hurricane model earns its place on the forecasting map, a new checklist for the Florida storm season, plus pressure from Congress halts the dismantling for now of a $386 million ocean observatory project. It’s all in this week’s Disaster Management Digest.
The Saharan Air Layer of dust from 2023. Courtesy, Jason Dunion NOAA/HRD
Saharan Dust Shear: Good news for Floridians and coastal Americans – just after torrential rains from Tropical Storm Arthur hit the Gulf Coast, Saharan dust from across the Atlantic is causing wind shear which should discourage hurricane formation through July. The dust is causing things to heat up stateside, dashing the hopes of South Floridians who are still facing wildfires in Miami-Dade area. Another weather phenomenon called the Madden Julian Oscillation (MJO) is in the mix, which often stirs up storm formation by causing air to rise, but given the circumstances, should be less impactful. National Weather Service and Colorado State University researchers are in agreement on the low hurricane likelihood but warn that “heat will remain a problem” as the warm winds cause the heat index to rise into the 105-to-110-degree range and evaporate clouds and most chances of rain. The next few weeks should be quiet, albeit swelteringly hot, in the Atlantic.
The accurate DeepMind forecasted track (in purple dash marks) of September 2025’s Hurricane Melissa. Courtesy NOAA
AI Hurricane Modeling: Artificial Intelligence has found its way into hurricane forecasting, and based on last year’s results, it was more accurate than any traditional method in predicting the path of Hurricane Melissa. The category 5 storm devastated Jamaica, and while other models flip-flopped on the tract and power of the storm, Google’s DeepMind insisted that the storm would develop into a Cat 5 and remain as such after landfall on the island. The same situation developed with the pattern of Hurricane Imelda last year – DeepMind was an extreme outlier in prediction accuracy. Such discrepancies in forecasting are at the heart of the debate: what are the AI models doing that traditional meteorologists are missing out on? The main difference is in methodology. Meteorologists are trying to use the most current data to predict the weather via atmospheric physics – the AI is poring over all available historical data and finding the best fit. Given last year’s success as an experimental model, this year the National Hurricane Center has partnered directly with Google DeepMind to develop a new AI hurricane forecast model that you’ll see labeled on tracking maps as “DeepMind.”
Hurricane Checklist: Before, During, and After: Speaking of getting ahead of the hurricanes, we at LMA see a lot of storm preparation guides, but I especially like this recent one from the Miami Herald. It covers not only preparation, but advice on what to do during and after the storm, including some reminders I haven’t seen in a while such as “Don’t use your generator during the storm, even if you lose power. Wait for the storm to pass before turning it on.” Now that we’re in hurricane season, we highly recommend you print this one, as well as the official Disaster Supply Kit Checklist from FloridaDisaster.org. We hope you’re well into executing my Lisa’s Lucky 7 Hurricane Season Prep, too!
Locations of planned descoping of data collection arrays. Courtesy, Ocean Observatories Initiative
Ocean Observatories Initiative Endangered: Over the last ten years, a network of more than 900 ocean sensors has tracked ocean circulation, marine ecosystems, climate change, and extreme weather under The Ocean Observatories Initiative. The project is federally funded to provide free data to over 500 scientific publications and was ready to run for another 15 to 20 years – that is, until the Trump Administration announced plans to dismantle the $386 million project under a broader retreat from climate-related issues. Last month, a group of Democratic senators and one Republican joined forces with two Democratic House committees and penned letters to the National Science Foundation (NSF) to reverse its decision to remove instruments from coastal waters by 2027, a decision which they contend, was made with no warning or scientific review. After one array of instruments was removed in Washington and Oregon, the widespread complaints reached the right ears, culminating in a reversal of the withdrawal, and NSF announcing plans to convene an expert panel to find the best path forward via a “thorough review and assessment” of the network.
LMA Newsletter of 7-6-26
