NASA scientists have developed a new artificial intelligence system designed to address one of the most persistent environmental and public health challenges affecting coastal waters: the detection and monitoring of harmful algae outbreaks. The research, recently published in the journal Earth and Space Science, demonstrates how AI can combine data from multiple satellites to identify harmful events in regions such as western Florida and Southern California. The innovation could significantly improve how scientists and local authorities detect and respond to algal blooms before they spread and cause major damage.
Harmful algae outbreaks are a growing concern because they can have severe impacts on ecosystems, public health, and coastal economies. In Florida, coastal areas such as Tampa Bay and Sarasota have struggled with recurring events caused by the species Karenia brevis. These outbreaks can kill fish and marine wildlife, contaminate beaches, and create health risks for people using coastal waters. On the U.S. West Coast, another species called Pseudo-nitzschia has produced toxins linked to the deaths of dolphins, sea lions, and other marine animals. In some cases, airborne toxins can even trigger respiratory problems in humans. Because of these consequences, monitoring algal blooms has become increasingly important for governments and environmental agencies.
Current monitoring methods rely heavily on field sampling and laboratory analysis. Scientists and health agencies often spend hours collecting water samples by boat before sending them to laboratories for testing. Results can take a day or longer, and repeated sampling may be needed as conditions change. This process becomes especially difficult because researchers often do not know where a bloom will emerge or how quickly it may spread. Organizations such as the National Oceanic and Atmospheric Administration (NOAA) already issue bloom forecasts, similar to weather predictions, but improving early detection remains a major goal.
NASA’s new AI system aims to strengthen these efforts by taking advantage of the extensive information already being collected by Earth-observing satellites. Satellites can detect numerous environmental signals associated with algal blooms, including changes in ocean color, pigment concentrations, and faint emissions of light produced during photosynthesis. For example, instruments aboard NASA’s PACE satellite can distinguish algae by characteristics such as size, shape, and pigment composition, while the TROPOMI instrument can identify subtle red fluorescence signatures associated with species like Karenia brevis.
Researchers faced a significant challenge in handling enormous quantities of satellite data collected from multiple missions and instruments. To overcome this, they developed a self-supervised machine learning system that can recognize relationships among different datasets without requiring manually labeled examples. The system learned patterns from satellite observations gathered in 2018 and 2019 and then compared those patterns with real-world measurements collected in the field.
Initial results showed that the AI system successfully detected and mapped harmful algal blooms, even in difficult coastal environments affected by sediment, vegetation, and runoff. NASA researchers believe that expanding the system with additional data from coastlines and lakes could eventually provide decision-makers with faster and more accurate environmental intelligence, helping industries ranging from tourism to aquaculture respond more effectively.
https://phys.org/news/2026-05-ai-tool-fuses-satellite-datasets.html