POSTER Session 4

MAPPING CYANOBACTERIA OCCURRENCE AND BLOOM RISK IN A TROPICAL LAKE

Cyanobacteria blooms threaten water quality in lakes and reservoirs worldwide. Existing remote sensing approaches for cyanobacteria focus on quantifying diagnostic pigments like phycocyanin or mapping surface accumulations by buoyant species in calm weather. However, these approaches are not fully complementary and do not provide intuitive understanding of water quality risks under all conditions. Optical Water Types (OWTs) offer a classification framework for inland and ocean waters to dynamically select suitable algorithms in evolving water conditions. Yet, the widely adopted OWT framework for lakes derives from a limited in-situ dataset, primarily from temperate regions, underrepresenting biogeochemical processes in complex tropical waters. Here, we extend an OWT framework to capture various states of cyanobacterial bloom in Lake Victoria, an optically complex tropical lake. Sentinel-3 Ocean and Land Colour Instrument (OLCI) images were used to extract spectra from 18 million pixels associated with cyanobacteria presence. K-means clustering was applied to explore intra-class variabilities, and similarities between candidate types were analyzed to design a Cyanobacteria Risk Index (CRI) leveraging information from various cyanobacteria classes. Our 25 candidate water types outperform the original 13 types, and the prevalence-based CRI offers an alternative risk assessment model, grouping risk into three categories for clearer interpretation. This study highlights the shortcomings of the current OWT framework for tropical waters, while our extended library and CRI capture bloom phases and transitional waters, improving cyanobacteria risk mapping. This approach enables proactive water management strategies and offers a baseline for future cyanobacteria risk assessments and management in tropical, optically complex waters.

*Davide Lomeo, King’s College London, United Kingdom, [email protected], https://orcid.org/0000-0002-9139-6834

Stefan Simis, Plymouth Marine Laboratory, United Kingdom, [email protected]

Emma Tebbs, King’s College London, United Kingdom, [email protected]

Nick Selmes, Plymouth Marine Laboratory, United Kingdom, [email protected]

Xiaohan Liu, Plymouth Marine Laboratory, United Kingdom, [email protected]

Mark Warren, Plymouth Marine Laboratory, United Kingdom, [email protected]

Code of Conduct
Privacy Policy