10 hours Ago
NEW ORLEANS — Artificial intelligence — or AI -- is everywhere nowadays. And that includes hurricane forecasting. It was a big topic at the National Hurricane Conference, which is going on this week in New Orleans.
KHOU 11 Chief Meteorologist David Paul is there and spoke to Dr. Philip Klotzbach with Colorado State University about the role AI can play in predicting what a storm may do. For years, meteorologists have been using traditional physics-based models for forecasting tropical weather.
That data comes through satellites, buoys, weather stations, aircraft and more. The speed and reliability of processing that data has improved over the years. But AI can take it a step further.
AI can take vast amounts of past weather data to essentially learn patterns, as opposed to solving physical equations. And it can do it much faster. RELATED: Small business owners: This resource is for you to get ready for hurricane season “That works really, really well on these shorter-term or when you have a ton of data," he said.
"So like on a day-to-day weather prediction and so they've already shown quite a bit of promise, especially for track prediction, for hurricanes and then a whole lot of other just broad forecasting of weather conditions.” As far as seasonal forecasting, Klotzbach said because there’s only 40 or 50 years of good data, you don’t necessarily need an AI model. “When we've used AI to develop our seasonal forecast model, effectively it gives us the predictors that we already selected, which is good,” he said.
“It's good to see that, but it's not really necessarily telling us something we didn't already know, per se.” Klotzbach said AI is here to stay, likely as complementary to traditional methods. You can watch Dr.
Klotzbach’s full interview streaming on KHOU 11+. Download on your TV and watch for free ! RELATED: Experts warn of energetic 2025 hurricane season with 17 named storms.
