AI Predicts Harmful Algal Blooms Using Satellite Data Off Portugal.

Ayman Bnoussaad, El Khalil Cherif, Ligia Pinto, Ramiro Neves, Alexandra D. Silva, Alexandre Bernardino· July 10, 2026 View original

Key takeaways

  • Satellite data can effectively predict harmful algal blooms using machine learning.
  • Ensemble tree-based models show strong performance in spatio-temporal HAB forecasting.
  • Seasonal, spatial, and lagged environmental factors are crucial predictors.
  • The framework supports operationally relevant early-warning systems for coastal regions.

Who benefits

AquacultureFisheriesTourismPublic HealthEnvironmental Monitoring

Summary

Researchers developed a machine learning framework using satellite data to predict harmful Pseudo-nitzschia algal blooms along the Portuguese coast, achieving moderate predictability with ensemble tree-based methods. The system identifies key environmental and biological factors influencing bloom occurrence, offering a new tool for early warning.

A new study introduces a machine learning system designed to forecast harmful algal blooms (HABs) caused by Pseudo-nitzschia diatoms along Portugal's Atlantic coast. This framework leverages exclusively satellite-derived predictors, such as sea surface temperature, chlorophyll-a, and plankton types, to predict bloom occurrences under realistic forecasting conditions. The models were developed using a decade of data from specific shellfish production zones, employing a stringent spatio-temporal cross-validation to ensure real-world applicability. Ensemble tree-based methods, particularly Random Forest and Extra Trees, demonstrated the strongest predictive power, with feature importance analysis highlighting seasonal patterns, spatial context, and lagged environmental conditions as primary drivers. This research provides a robust, operationally relevant early-warning system for HABs in eastern boundary upwelling coasts, enhancing environmental monitoring and public health safety.

Why it matters

This research offers a critical advancement for environmental monitoring and public health, enabling earlier detection and mitigation of harmful algal blooms that impact coastal ecosystems and industries.

How to implement this in your domain

  1. 1Integrate satellite data streams into existing environmental monitoring platforms.
  2. 2Develop or adapt machine learning models for specific regional HAB prediction.
  3. 3Establish protocols for disseminating early warnings to affected industries and communities.
  4. 4Collaborate with research institutions to refine predictive models and incorporate new data sources.

Original post by Ayman Bnoussaad, El Khalil Cherif, Ligia Pinto, Ramiro Neves, Alexandra D. Silva, Alexandre Bernardino

"arXiv:2607.07834v1 Announce Type: new Abstract: Pseudo-nitzschia diatoms pose recurrent risks to coastal ecosystems and shellfish harvesting along the Portuguese Atlantic coast. Here we develop and evaluate a spatio-temporal machine-learning framework to predict harmful algal blo…"

View on X

Originally posted by Ayman Bnoussaad, El Khalil Cherif, Ligia Pinto, Ramiro Neves, Alexandra D. Silva, Alexandre Bernardino on X · view source

Want to go deeper?

Turn these trends into skills with Learnijoy's hands-on AI & tech courses.

Explore courses

More in AI Research