Systematic analysis of the evolution of modelling approaches applied in algal bloom forecasting

Algal blooms occurrence has increased in aquatic ecosystems in recent years. They can be influenced by wastewater from settlements and agro-industry. This type of water is characterised by high nutrient concentrations, which can cause eutrophication and favour algal blooms. This event impacts the aquatic environment by reducing dissolved oxygen levels and causing water pollution through the possible formation of toxins. According to the Food and Agriculture Organisation of the United Nations (FAO), several approaches have been developed to understand its nature and predict algal blooms. This study focuses on analysing predictive modelling of algal blooms using bibliometrics and a systematic review to identify the main model types that enable understanding of microalgae dynamics. Our approach consists of three phases: i) definition of search criteria and database collection; ii) analysis of conceptual evolution and structure; and iii) systematic review and identification of model types and research trends. The study demonstrated that hybrid approaches (combining mechanistic models with AI algorithms) overcame traditional limitations in interpretability and scalability. In addition, two lines of future research for understanding algal dynamics were identified: (1) expanding studies on tropical and transitional environments and (2) developing hybrid models to address climate change and water security. Our research provides an optimal framework for selecting models based on environmental type, target, and data availability, offering a practical guide for analysing the dynamics of harmful algal blooms (HABs) in a context of global change.