From forecast to operational decision-making
Operational ocean forecasting is about delivering up-to-date information on the expected development of ocean conditions in a format that can be used directly for planning and operational management. We develop systems that automatically retrieve weather and ocean forecasts from sources including the Norwegian Meteorological Institute, ECMWF, NOAA and Copernicus Marine Service, and refine them for specific coastal, fjord or operational areas.
Through FjordLab, we are developing an open and modular framework that connects data sources, forecasting models, observations and user-oriented services through standardised interfaces. Results can be delivered as maps, time series, probabilistic forecasts, threshold-based alerts or machine-readable data that can be integrated directly into customers' own systems.
Situational awareness for defence and underwater operations
For defence and underwater technology applications, operational ocean forecasts can describe conditions that influence sensors, acoustics, communication and underwater mobility. Forecasts of temperature and salinity at different depths can be combined with pressure data to generate sound velocity profiles and three-dimensional sound speed fields. These can then be used as input to models of acoustic propagation, sonar performance and underwater communications.
Forecasts of current speed and direction throughout the water column can be combined with vehicle performance characteristics, energy requirements and operational constraints. This provides a basis for calculating routes, depths and operational time windows that reduce energy consumption, improve station keeping and increase mission success rates.
Fisheries: more targeted and efficient search operations
For the fisheries industry, forecasts of temperature, fronts, currents and plankton can be used to identify areas where favourable environmental and feeding conditions are expected. Combined with catch data and acoustic measurements from vessels and autonomous platforms, these forecasts can support the development of dynamic search strategies in which routes and search patterns are updated in response to changes in the marine environment. This can reduce search time, energy consumption and costs, while supporting future autonomous fishing operations.
Aquaculture: early warning and risk management
For aquaculture, operational forecasts can track developments in oxygen concentration, temperature, currents, stratification and water exchange at farming sites. These variables influence fish welfare, the dispersion and sedimentation of waste products, and operations such as feeding, handling, delousing and well-boat visits.
Forecasts can be compared against site-specific threshold values and presented as expected risk levels for the coming hours and days.
Environment, emissions and ecosystem impacts
Operational ocean forecasting can be used to monitor how emissions, nutrients and particles are transported and interact with local ecosystem conditions. This is relevant for industry, regulators and emergency response organisations operating in fjords and coastal areas with multiple sources of environmental pressure.
The systems can calculate how inputs from local industry, rivers or aquaculture facilities are transported over the coming days and how they interact with oxygen conditions, algal growth or environmentally sensitive areas. During incidents, forecasts can be updated rapidly and used to assess areas of impact and prioritise monitoring activities or mitigation measures.
Rapid short-term forecasting using artificial intelligence
When decisions must be made within minutes, AI-based models can complement conventional forecasting systems. We can train fast emulators and nowcasting models using high-resolution simulations and continuously updated observations. These models can then predict future developments with very low latency and high temporal resolution.
Such solutions can provide updates between regular model runs, perform rapid scenario analyses or operate locally on board vessels, autonomous vehicles and control systems. The physics-based model remains the reference and training foundation, while AI makes predictive capabilities available when rapid response times are critical.