Objectives of the service
Harmful algal blooms (HABs) are a recurring and costly threat to salmon farming in Chile, causing sudden fish mortality, farm closures, and losses that can reach millions of dollars per event. Today, farmers rely on fragmented monitoring systems based on local sensors and periodic sampling, offering limited anticipation and little time to act.
PREDIFAN delivers a new generation early warning and forecasting service designed specifically for aquaculture operations. It provides reliable, timely, and actionable insights that enable farmers to anticipate HAB events and take preventive measures, such as adjusting feeding strategies, activating aeration systems, or relocating cages before impacts occur.
The service combines satellite observations, in-situ measurements, and advanced hydrodynamic and biological models within a single, user-friendly platform. By transforming scattered data into clear operational guidance, PREDIFAN supports faster and more confident decision-making.
The project aims to reduce fish mortality, improve animal welfare, and lower operational risks and costs. It also helps farms meet environmental requirements while strengthening the long-term resilience and sustainability of the aquaculture industry
Users and their needs
PREDIFAN targets the salmon aquaculture sector in Chile, particularly farming operations in the southern regions where harmful algal blooms (HABs) represent significant environmental and economic risks. Core users include farm managers, production managers, and environmental monitoring teams responsible for daily operations and risk mitigation. Regulatory authorities are also stakeholders, requiring reliable data to ensure compliance and assess ecosystem impacts.
These users need timely, actionable, and site-specific insights to anticipate HAB events and make informed operational decisions. However, they currently face fragmented data sources, limited predictive tools, and high uncertainty in risk assessment.
Key user needs and challenges:
-
Early and reliable warning of HAB events
-
High spatial resolution at farm level and in Farming Concession Areas (ACS)
-
Robust, validated data and risk indicators
-
Intuitive, user-friendly dashboards for non-experts
-
Continuous service availability
The main challenge for PREDIFAN is to transform complex, multi-source environmental data into accurate, operational, and easy-to-use decision-support tools. Building user trust in predictions and ensuring seamless integration into existing workflows will be critical for adoption and long-term impact.
Service/ system concept
PREDIFAN delivers a user-oriented digital service that transforms complex environmental data into clear, actionable insights for aquaculture operators. The platform provides real-time monitoring and short-term forecasts (72 hours) of harmful algal bloom (HAB) risks at farm level through an intuitive dashboard.
The system combines multiple data sources, including satellite observations to detect large-scale phytoplankton activity, in situ measurements from farms for local accuracy, and hydrodynamic models to simulate bloom transport and evolution. These inputs are processed within a central system and translated into simple, site-specific risk indicators.
Users can access daily updated maps showing current conditions, bloom dynamics, and short-term forecasts. A built-in alert system automatically notifies operators when critical thresholds are exceeded, enabling rapid and proactive responses such as adjusting feeding or deploying mitigation measures.
By integrating diverse data into a single, easy-to-use platform, PREDIFAN shifts from reactive monitoring to predictive decision support. This improves situational awareness, reduces operational risks, and supports more efficient and resilient aquaculture practices.
Space Added Value
PREDIFAN leverages satellite Earth observation data from the Copernicus Sentinel-2 and Sentinel-3 missions to continuously monitor ocean conditions and derive key indicators such as chlorophyll-a concentration and phytoplankton functional types. These space-based datasets are enriched with in-situ measurements from aquaculture sites to ensure high local accuracy.
The added value of space assets lies in their unique capacity to provide frequent, large-scale, and consistent observations across coastal and offshore areas that are otherwise sparsely monitored. Unlike conventional methods—limited to localized, reactive sampling—satellite data enables early detection of harmful algal bloom formation offshore and continuous tracking of their evolution.
By combining Earth observation data with hydrodynamic transport models and advanced machine learning, PREDIFAN delivers predictive capabilities on bloom development and transport. This combined use of space assets and modelling significantly improves anticipation and reduces uncertainty.
Compared to existing approaches relying primarily on in-situ measurements, PREDIFAN enables a proactive, forecast-based service, supporting timely decision-making and risk mitigation. The synergy between space-based observations and modelling thus provides a clear operational and economic advantage compared to traditional monitoring systems that rely solely on local measurements and lack forecasting capability.
Current Status
The PREDIFAN project has completed its system definition phase, with the Requirements Document and System Verification Document established and aligned with ESA expectations. User needs were identified and consolidated during AquaSur 2026 event (Chile), through direct exchanges with salmon farming stakeholders, providing a structured and operational basis for system design.
The system architecture, data flows, and verification strategy have been defined, including a complete verification matrix ensuring traceability between requirements and planned tests. Key components such as data ingestion, modelling approaches, and alerting logic have been specified but not yet implemented.
Current activities focus on finalising documentation and preparing for the Critical Design Review (CDR).
Following CDR approval, the next phase will consist of system development, integration of data sources (in situ, satellite, oceanographic), and subsequent validation activities with pilot users.
Prime Contractor(s)
Subcontractor(s)
