FSU’s Bingqing Liu explains how AI could turn red-tide tracking into earlier warnings

AI could help anticipate toxic red tides by combining satellite observations with ocean models and environmental data, Florida State University researcher Bingqing Liu said in an October 27 Q&A. Earlier exposure warnings are the promise; the underlying evidence comes from an August study that traced changes in algae after Hurricane Milton.

Liu describes AI looking for patterns that precede blooms, including changes in phytoplankton communities, temperature, salinity, currents and nutrients. She says those signals can support earlier detection and short-term forecasts, giving people vulnerable to respiratory or neurological symptoms time to take precautions.

Separating the algae in ocean color

The Scientific Reports study, published August 27, examined a Karenia brevis bloom after the 2024 hurricane on the West Florida Shelf. Researchers used PhytoUnmix, a deep-learning system that separates detailed light measurements from NASA’s PACE satellite into information about different phytoplankton groups—microscopic organisms in the water.

That distinction is central to Liu’s explanation. She says PACE’s detailed ocean-color observations let scientists distinguish phytoplankton types, including Karenia brevis, the organism responsible for Florida’s toxic red tides. The goal is to identify harmful bloom development rather than measure only the overall abundance of algae.

While we cannot prevent red tide itself, we can significantly improve early warning and reduce its impacts on communities and ecosystems.

Bingqing Liu, in Florida State University’s October 27 Q&A

Milton’s uneven biological aftermath

The hurricane’s aftermath did not produce a uniform response. Offshore, wind-driven upwelling was associated with a higher relative abundance of diatoms, one phytoplankton group. Near shore, northerly winds, water transport and enhanced upwelling were associated with the Karenia brevis bloom.

Combined observations from PACE and the Surface Water and Ocean Topography satellite showed a rapid nearshore rise in dinoflagellates. Diatoms and chlorophytes declined, while cryptophytes increased. The authors interpret the shifts as likely reflecting both physical stirring and biological responses, with freshwater inputs increasing nutrient availability.

From bloom detection to health protection

The health concern extends beyond contact with seawater. Liu explains that waves can break open algal cells and release brevetoxins into the air. Inhaling them can cause respiratory irritation and potentially neurological symptoms such as headaches or brain fog, she says. Toxins can also accumulate in shellfish and pose risks when eaten.

The August study supplies evidence about bloom composition and storm-related ocean conditions. It examines an environmental event, not whether an AI warning system reduced illness. Turning those observations into health protection is the application Liu describes.