Reliability of oil spill modeling in Arctic waters for response planning and risk assessments depends on the accuracy of winds, currents, and ice data (cover and drift) used as input. We compared predicted transport in ice, using ice and ocean model results as input, with observed drifter trajectories in the Beaufort Sea and an experimental oil release in the Barents Sea. The ice models varied in ice rheology algorithms used (i.e., ElasticViscous-Plastic, presently used in climate models, versus a new Elasto-Brittle approach in pack ice) and the time averaging of their outputs, which were provided as input to oil spill models. Evaluations of model performance (skill) against drifters showed improvement using Elasto-Brittle instead of Elastic-Viscous-Plastic rheology. However, model skill was degraded by time-averaging of ocean and ice model vectors before input to the oil spill model. While the accuracy of individual oil model trajectories projected weeks to months into the future is expected to be low, in the event of a spill, forecasts could be updated frequently with satellite and other observations to improve reliability. Comparisons of modeled trajectories with drifters verified that use of the ice-ocean models for ensemble modeling as part of risk assessments is reliable.Key words: trajectory model, ice drifter, ice model, Arctic spill response, oil weathering in ice.Résumé : La fiabilité de la modélisation de déversement de pétrole dans les eaux arctiques en vue d'un plan d'intervention et d'évaluations de risque dépend de l'exactitude des données sur les vents, les courants et les glaces (la couverture et la dérive) utilisées comme données d'entrée. Nous avons comparé le transport prévu dans les glaces, utilisant les résultats de modèles des glaces et océanique comme données d'entrée, avec des trajectoires de dérive observées dans la mer de Beaufort et un rejet de pétrole expérimental dans la mer de Barents. Les modèles des glaces variaient selon les algorithmes de rhéologie de glace utilisés (c'est-à-dire, plastique-visqueux-élastique, actuellement utilisé dans les modèles climatiques, contre une nouvelle approche élasto-fragile dans la banquise) et la détermination des moyennes de leurs résultats, fournis comme données d'entrée pour les modèles de déversement de pétrole. Les évaluations de la performance (capacité) du modèle par rapport aux dérives ont présenté une amélioration lorsqu'on utilisait la rhéologie élasto-fragile au lieu de plastique-visqueux-élastique. Cependant, la capacité du modèle s'est dégradée à cause de la détermination des moyennes de vecteurs des modèles océaniques et des glaces avant que ceux-ci ne soient introduits dans le modèle de déversement de pétrole. Tandis qu'on s'attend à ce que l'exactitude des trajectoires projetées par des modèles pétroliers particuliers des semaines à des mois à l'avenir ne soit pas grande, dans le cas d'un déversement, les prévisions pourraient fréquemment être mises à jour à l'aide d'observations par satellite et Mots-clés : modèle à tr...

No abstract

The California Department of Fish and Game Office of Spill Prevention and Response (CA OSPR) is utilizing oil-spill fate and transport modeling to develop the time and spatial scales, and equipment needs, for a formal Dispersed Oil Monitoring Plan (DOMP). When fully implemented, the DOMP will aid in documenting hydrocarbon concentrations in the water column, potentially exposed organisms (zooplankton), and the impacts of entrained oil and dissolved hydrocarbons with and without dispersant applications. Fluorescein dye studies off San Diego, California (USA) have been completed to test the operational framework for repeated sampling of dispersed oil plumes as outlined in the DOMP, to allow evaluation of high-frequency radar (HF-Radar) for providing surface current input data to oil spill models, and to provide verification of model-predicted movement of subsurface oil (dye) by comparison with drogue movement and measured dye concentrations over three dimensions and time. Aerial photodocumentation, subsurface drogues, dye transport, and HF-Radar were used to measure near-surface current fields at varying depths. High-resolution subsurface dye-plume structure was mapped using an in situ GPS-coupled towed fluorometer equipped with pressure sensors to provide dye concentration data as a function of time, position, and depth. In addition, data from the more traditional Special Monitoring of Applied Response Technology (SMART) protocols utilized by the U.S. Coast Guard (USCG) were compared with the in situ towed-fluorometer measurements, and conventional CTD data were collected to determine the mixed layer depth, an important variable in monitoring dispersion of oil in the water column. As a result of these efforts, significant progress has been made on developing and testing sampling protocols for the DOMP, and nearly continuous and synoptic data have been obtained from seven cruises conducted over a 12-month period. These data sets (available on-line through the Coastal Response Research Center (CRRC) website: http://www.crrc.unh.edu/) are being analyzed and integrated to support oil spill model development and verification with direct applicability to spill response decision making, net environmental benefit analysis, natural resource damage assessments, and educating the spill community and public.