SINTEF Ocean has for years developed the OSCAR — a software for predicting the spread and fate of oil spill at sea which is used operationally in emergency situations and for risk assessment related to contingency planning.
At the same time, SINTEF Digital has developed GPU Ocean — a GPU-accelerated simplified ocean model for quantifying short-term uncertainties of ocean currents. The model also is well-suited for running ensembles of high-resolution simulations nested within complex operational ocean circulation models.
In this project, we will develop a GPU-accelerated Lagrangian stochastic transport model for oil with an online one-way coupling to the existing simplified ocean model GPU Ocean. The goal is to run large ensembles to obtain short-term uncertainty estimates in events such as near-shore accidents. Through this work, we will combine SINTEF Digital's expertise in GPU-accelerated computing with SINTEF Ocean's domain-expertise within oil spill contingency planning.