Introduction: the Earth Observation for Sustainable Aquaculture (EO4SA) project
Aquaculture offers an additional supply of human food resources as well as an alternative to overfishing. Accordingly, aquaculture is a means to meet the growing global demand for (sea)food and a low carbon footprint protein source, but the sustainability of farming activities is impacted by several key factors, of which four problem areas are being addressed in the four EO4SA Use Cases:
Predicting Risks of Salmon Lice Infestation in Norway.
Forecasting Harmful Algal Blooms (HABs) Impacting Shellfish Farming in Norway.
Optimising the Location of Sustainable Shellfish Farming and Tourism in Galicia, Spain.
Mapping Aquaculture Structures and Use of Marine Resources in Palawan, Philippines
These factors or the environmental conditions associated with them are possible to monitor directly or indirectly by Earth Observation (EO). Therefore, we have developed innovative EO high-level indicators for supporting the sustainable development, regulations and management of aquaculture farming in coastal and offshore waters. This presentation will provide a summary of the project results to date and to collect feedback from the public and private user community for future development of satellite products.
Predicting Risks of Salmon Lice Infestation in Norway
Risk forecasting of salmon lice infestation is a promising EO-based application, leveraging large in situ datasets (>690,000 records) combined with AI models. The project has taken the ocean variables sea surface temperature and salinity, together with the in situ fish farm data, to develop a model to forecast salmon lice outbreaks. A weekly forecast for all Norwegian fish farms is now publicly available, which gives the probability of crossing the lice per fish thresholds that are specified by Norwegian regulations.
Forecasting Algal Toxin Contamination Impacting Shellfish Farming in Norway.
Key environmental conditions from a coastal ocean model, together with a machine learning model to estimate the probability of algal toxins being detected in blue mussels seven days ahead. The forecasted thresholds are above 20 µg/kg for diarrheic shellfish toxins – DST and above 120 µg/kg for paralytic shellfish toxins – PST. The method builds on and extends the published AI model for Lyngen fjord (Silva et al. 2023), now recalibrated using datasets from the entire Norwegian coast (Silva et al. 2024). It is intended that this forecast can help mussel farms in planning mussel harvesting.
Optimising the Location of Sustainable Shellfish Farming and Tourism in Galicia, Spain
The National Park Illas Atl��nticas de Galicia (Spain) located at the entrance of the R��as Baixas, four large coastal embayments in the western Galician coast, hosts biodiversity-rich ecosystems and is fully protected, promoting eco���sustainable tourism. Visitor numbers are increasing (451,000 in summer 2024, +5% year���on���year), supporting 26 certified sustainable tourism companies. Although aquaculture is currently prohibited, growing tourism raises interest in identifying surrounding areas for potential touristic expansion or future protection. This case study assessed areas with high mussel growth potential to support tourism-related activities near the park and inform long-term conservation planning.
Mapping Aquaculture Structures and Use of Marine Resources in Palawan, Philippines
Aquaculture infrastructure is mapped using active and passive multispectral EO data to detect structures such as mussel long���lines and fish cages. Building on previous work using a Sentinel���1/2–based method developed for Malampaya Sound (Kurekin et al. 2022), the approach has been extended with artificial intelligence to Puerto Princesa Bay, Palawan (Phillipines). A convolutional neural network previously proven in Palawan waters has been retrained and adapted to meet Early Adopters' needs. Historical very���high���resolution ESA Third���Party Mission imageries were used for training and validation of the detection algorithms.
Acknowledgment
Project Earth Observations for Sustainable Aquaculture (EO4SA), supported by the European Space Agency contract n 4000146872/24/I-EF.
References
Kurekin, Andrey A., Peter I. Miller, Arlene L. Avillanosa, and Joel D. C. Sumeldan. 2022. Monitoring of Coastal Aquaculture Sites in the Philippines through Automated Time Series Analysis of Sentinel-1 SAR Images. Remote Sensing 14 (12): 2862. https://doi.org/10.3390/rs14122862.
Silva, Edson, Julien Brajard, Fran��ois Counillon, Lasse H. Pettersson, and Lars Naustvoll. 2024. Probabilistic Models for Harmful Algae: Application to the Norwegian Coast. Environmental Data Science 3: e12. https://doi.org/10.1017/eds.2024.11.
Silva, Edson, Fran��ois Counillon, Julien Brajard, Lasse H. Pettersson, and Lars Naustvoll. 2023. Forecasting Harmful Algae Blooms: Application to Dinophysis Acuminata in Northern Norway. Harmful Algae 126 (July): 102442. https://doi.org/10.1016/j.hal.2023.102442.