A better understanding of how the different input and model parameters affect the oil spill model predictions will help to improve the model performance and thus better assist emergency response and provide insights on ecosystem effects. This research studies the influences of spatial and temporal resolutions of hydrodynamics forcing as well as oil spill model parameters, such as number of particles and grid size, on the predicted surface, water column, and shore impact areas. A case study using the OSCAR oil spill models was conducted to investigate these effects for a hypothetical spill on Scotian shelf. The model was forced with the daily HYCOM, daily NEMO, and hourly NEMO hydrodynamic forcing and configured with three different number of particles and three grid settings. The results show that all these parameters could affect the prediction at various degrees. Special attention should be paid to certain parameters depending the study objectives (e.g. surface or shoreline impacts).
The recent Deepwater Horizon disaster has shown that oil spills in the deep-water environment present extremely difficult response logistics; hence, it is vital that a reliable oil spill model be available to predict the trajectory of oil and delineate the allocation of limited response resources. To date, there are a number of models have been developed for this purpose, and examples of these models were given by Socolofsky et al. (1995): the SINTEF Oil Spill Contingency and Response (OSCAR) model (Reed et al., 1996b); The National Energy Technology Laboratory (NETL) Blowout and Spill Occurrence Model (BLOSOM) (NETL, 2017); The MIKE by DHI Oil Spill (OS) module, with integrated near-field plume model and coupled Lagrangian and Eulerian model for tracking of dispersed and dissolved oil in the farfield (DHI, 2015); RPS-ASA's OILMAP model which includes OILMAPDeep model as the integrated near-field plume model (ASA, 2017).
The intercomparison of these models with similar input information by Scolofsky et al. (2015) has indicated while the models agreed reasonably well in some respects but they differed in other respects by an order of magnitude or more. Furthermore, the oil spills models use particle tracking approach to trace parcels of oil moved by ocean currents and wind, the resolution of these inputs could have a significant effect on the predicted trajectory. To employ these models for more reliable prediction, it is important to know the sensitivity of model to different user defined parameters and inputs. Although there are sensitivity studies of oil spill models to selected model parameters such as particle number, spreading coefficient etc., there lacks information on the effects of external forcing, such as spatial and temporal resolutions of hydrodynamic models on the model predictions. The fill this information gap, this study aims to study the effects of both oil spill model inputs and user defined parameters on the trajectory and mass balance of oil from subsurface blowouts.
