Harmful algal blooms (HABs) occur frequently in Guangxi coastal waters, posing potential risks to marine ecosystems, aquaculture, and critical coastal infrastructure. However, conventional HAB monitoring methods often rely on a single chlorophyll-a threshold and may not adequately account for site-specific background conditions or coordinated changes in multiple water-quality parameters.
This study developed a multiparameter scheme for classifying HAB risk levels using hourly observations from five coastal buoy stations during 2020–2024, together with meteorological, tidal, and historical HAB records. Chlorophyll-a and dissolved oxygen were selected as the primary indicators. Quantile-based statistics, standard-deviation thresholds, and time-lag analyses were combined to characterize station-specific background variability and identify HAB-related anomalies. The scheme integrates anomaly magnitude, persistence, and coordinated responses between indicators, thereby reducing the sensitivity of single fixed-threshold methods to local background differences and short-term fluctuations.
Retrospective evaluation against historical HAB records showed that the proposed scheme achieved a probability of detection (POD) of 95.8%. Compared with a single chlorophyll-a threshold, the multiparameter scheme more comprehensively characterized coordinated biogeochemical changes during bloom initiation, development, and decline. It provides an objective and quantitative basis for HAB risk classification, continuous anomaly tracking, and operational monitoring and early warning in Guangxi coastal waters.