Model-based reconstruction of the Ulysse-Virginia oil spill, October–November 2018. (May 2020)

• Record Type: Journal Article
• Authors: Liubartseva, S.; Smaoui, M.; Coppini, G.; Gonzalez, G.; Lecci, R.; Cretì, S.; Federico, I.
• Is Part Of: Marine pollution bulletin. Volume 154(2020)
• Publication Date: 2020-05
• Subjects: Oil spill emergency -- REMPEC-MONGOOS collaboration -- CMEMS -- MEDSLIK-II -- Deterministic and stochastic forecasts -- Liguro-Provençal basin; Marine pollution -- Periodicals; Marine Biology -- Periodicals; Water Pollution -- Periodicals; Mer -- Pollution -- Périodiques; Publications périodiques; Pollution des mers; Lutte antipollution; Electronic journals; 363.7394
• DOI: 10.1016/j.marpolbul.2020.111002
• Languages: English
• ISSNs: 0025-326X

Abstract

Abstract: In response to the oil spill caused by the collision between the Ro-Ro ship Ulysse and CSL Virginia on 7 th October 2018, the Lagrangian oil spill model MEDSLIK-II was utilized to predict spill transport and fate. Oil drift was forced by the high-resolution sea circulation provided by CMEMS and the ECMWF wind. Successive model runs were restarted 5 times with the distinct overflight- and satellite-derived observations provided by REMPEC and the Copernicus Sentinel-1 mission. The results were verified based on the ability to predict the first oil-beaching event that happened near Saint-Tropez (France) in the early afternoon of 16 th October. Despite the general consistency among the runs, only the last initialization was able to forecast the oil beaching. Stochastic MEDSLIK-II simulations forced by the historical meteo-oceanographic datasets 2014–2018 revealed that the coastlines between Cap Lardier and the Gulf of Saint-Tropez were among the most impacted areas. Highlights: Ulysse-Virginia oil spill was modeled using various observations. REMPEC-MONGOOS operational collaboration improved quality of oil drift forecast. MEDSLIK-II kept a 16-day period of reliable predictions in deterministic mode. Stochastic MEDSLIK-II runs were useful to predict which areas were most affected. Both accuracy of model and quality of observations should be improved.