Tag
This study develops a machine learning framework using satellite-derived environmental and biological predictors to forecast Pseudo-nitzschia harmful algal blooms along the Portuguese coast, achieving up to 0.77 ROC-AUC with Extra Trees models under realistic constraints.