Evaluation of Oil Spill Modeling in Ice Against In Situ Drifter Data from the Beaufort Sea

Abstract

ABSTRACT While coupled ice-ocean models provide reliable hindcasts and large-scale predictions of ice conditions and movements in the Arctic, to date, operational models have not been implemented with sufficient spatial resolution or skill to define sea ice characteristics and dynamics needed for high resolution oil spill trajectory forecast modeling. Recently (2015) Nansen Environmental and Remote Sensing Centre (NERSC) researchers updated their modeling approach and rheology used for pack ice. They found that using the newly developed Elasto-Brittle (EB) model showed significant improvement in performance over the present Elastic-Viscous-Plastic (EVP) modeling approach used in the operational forecast and reanalysis versions of their TOPAZ4 coupled ice-ocean model. NERSC also integrated a wave-in-ice model (WIM) into a newly updated version of TOPAZ, to characterize waves in the Marginal Ice Zone (MIZ). RPS ASA’s oil transport and fate models OILMAP and SIMAP (OIL/Spill Impact Model Application Package) were updated, integrating the NERSC ice modeling products for use in transport and oil weathering algorithms. Oil trajectory model simulations, using the existing publically-available TOPAZ4 and updated ice model products, were compared with available in situ drifter data for the Beaufort Sea from the International Arctic Buoy Programme (IABP). The goal was to evaluate model performance (skill) against drifters that were trapped in the pack ice where the EB/EVP rheology applies. The comparisons show that model-based trajectories increasingly diverged from observations over days and weeks due to cumulative errors. The model using EB rheology more closely agreed with the IABP observations than TOPAZ with EVP, and the updated TOPAZ showed improved model performance over TOPAZ4. However, model skill was degraded by time-averaging of ocean and ice model vectors before input to the oil spill model. Demonstrated improvement of oil-in-ice spill modeling would help meet the needs for Arctic oil spill response in the coming decades. While the accuracy of individual oil model trajectories projected weeks to months into the future would be expected to be low, in the event of a spill, forecasts could be updated frequently (on a time scale of hours to days) with satellite information, aircraft observations, drifter data, and other observations to improve reliability. The overall transport patterns and results of an ensemble of trajectories would provide useful information for planning and risk assessments based on typical current and ice movement patterns.