Oil spill model uncertainty quantification using an atmospheric ensemble

Abstract

Abstract. We investigate the impact of atmospheric forcing uncertainties on the prediction of the dispersion of pollutants in the marine environment. Ensemble simulations consisting of 50 members were carried out using the ECMWF ensemble prediction system and the oil spill model MEDSLIK-II in the Aegean Sea. A deterministic control run using the unperturbed wind of the ECMWF high-resolution system served as reference for the oil spill prediction. We considered the oil spill rates and duration to be similar to major accidents of the past (e.g., the Prestige case) and we performed simulations for different seasons and oil spill types. Oil spill performance metrics and indices were introduced in the context of probabilistic hazard assessment. Results suggest that oil spill model uncertainties were sensitive to the atmospheric forcing uncertainties, especially to phase differences in the intensity and direction of the wind among members. An oil spill ensemble prediction system based on model uncertainty of the atmospheric forcing, shows great potential for predicting pathways of oil spill transport alongside a deterministic simulation, increasing the reliability of the model prediction and providing important information for the control and mitigation strategies in the event of an oil spill accident.

Assessing ocean ensemble drift predictions by comparison with observed oil slicks

“…We acknowledge that uncertainties in the wind field are the main sources of variability in large and persistent oil slick scenarios (e.g. Li et al, 2019;Kampouris et al, 2021), but we have evidences that it is overridden by the ocean currents on short scales.…”

“…Less attention has been paid to the uncertainties in ocean models compared to wind forecasts due to its perceived secondary role in marine pollutant modelling (Li et al, 2019;Kampouris et al, 2021), though Jones et al (2016) showed that the oil slick drift modeling was improved when ocean current data retrieved from drogued buoys were included in the drift modelling. Additionally, other studies have also shown that poorly resolved mesoscale features (Jorda et al, 2007) and inertial oscillations (Brekke et al, 2021) impact the modeled drift of slicks, thus misrepresenting the predicted trajectory if not taken into account.…”

Thyroid Hormone Levels in Mothers and Cord Blood at Delivery in Crude Oil Producing Community in Delta State, Nigeria

“…14, 25 Low maternal tetraiodothyronine (T4) concentrations during gestation are associated with the impaired cognitive development of offspring as indicated by decreased Fagans or Bayleys Scores. 26 This study, therefore, is of public health importance and may provide valuable data that could be used in the formulation of reproductive health policies aimed at reducing the effects of neurotoxicants on infants. Data from this study showed significant changes in thyroid hormone concentrations, association with BPDE as a marker of PAHs during pregnancy and effects on the infants.…”

Variance-based sensitivity analysis of oil spill predictions in the Red Sea region

“…Locations 1 and 2 display a nonmonotonic change in the total sensitivity, with the wind coefficient having the largest influence. This outcome is unsurprising because the shear stresses imposed by the surface winds stretch the oil slick along the main wind direction, directly distorting its shape (Kampouris et al, 2021). However, the effects of the IFT and viscosity are generally not negligible, as observed from the results for Locations 1 and 2.…”