The DeepWater Horizon Oil Slick: High Resolution Model Simulations of River Front Effects, Initialized and Verified by Satellite Observations

Abstract

The effect of river fronts on oil slick transport has been demonstrated using high resolution forcing models and a fully fledged oil drift model, OpenOil. The model system is used to simulate the 2010 DeepWater Horizon oil spill. Metocean forcing data are taken from the GoM-HYCOM 1/50° ocean model with realistic river input and ECMWF global forecast products of wind and wave parameters with 1/8° resolution. The simulations are initialized from satellite observations of the surface oil patch. OpenOil includes most of the relevant processes, such as emulsification, evaporation, wave entrainment, stranding and droplet formation. The model takes account of the actual oil type and properties, using the ADIOS oil weathering database of NOAA. The effect of using a newly developed parameterization for oil droplet size distribution is studied and compared to a traditional algorithm. Although the algorithms provide different distributions for a single wave breaking event, it is found that the net difference after long simulations is negligible, indicating that the outcome is robust regarding the choice of parameterization. That indicates that the wave entrainment, vertical mixing and re-surfacing mechanisms that are part of OpenOil are more important for determining the final droplet size spectrum than the spectrum prescribed for individual wave breaking events. In both cases, the size of the droplets controls how much oil is present at the surface and hence are subject to wind and Stokes drift. The effect of removing river outflow in the ocean model is investigated in order to showcase effects of river induced fronts on oil spreading. A consistent effect on the amount and location of stranded oil is found, and considerable impact of river induced fronts is seen on the location of the surface oil patch. During a case with large river outflow (May 20-27, 2010), the total amount of stranded oil is reduced by about 50% in the simulation with no river input. The results compares well with satellite observations of the surface oil patch.

Oil Spill Modeling: A Critical Review on Current Trends, Perspectives, and Challenges

“…The outputs demonstrate that an accurate description of fate and transport processes, in particular vertical mixing and oil weathering, is needed to represent the horizontal spreading of the oil spill [112]. This study found that the difference between two different oil droplet size distribution is negligible, demonstrating that the effect is strong in terms of the configuration option [262]. On the other hand, according to Röhrs et al [112] and the study of OPENOIL model, small dispersed droplets should not be ignored in oil spill modeling as they can appear as resurfaced oil.…”

“…OpenOil (Figure 4.) has been implemented operationally in Norway, as an oil spill contingency and search and rescue model [261,262] and for drifter and oil slick observations in the North Sea [112,263]. MET Norway uses in-house high-resolution ocean circulation and meteorological models as forcing for providing operational oil spill forecasts with OpenOil.…”

“…This model integrates algorithms with several physical processes, such as wave entrainment of oil [126], vertical mixing by virtue of oceanic turbulence [126,141], resurfacing of oil on account of buoyancy [111], and emulsification of oil properties [25,126]. Resurfacing is parameterized based on oil density and droplet size by means of Stokes Law, and for this reason the model's physics are very sensitive to the specification of the oil droplet size distribution [261,262]. In OpenOil, the oil properties are obtained from the ADIOS Oil Library.…”

“…Furthermore, the OpenOil model has recently indicated a great agreement with satellite observations of the DWH oil slick [261,262] for two seven-day simulations forced via the metocean results from the high resolution GoM-HYCOM 1/50 (in the GoM), FKEYS-HYCOM 1/100, and ECMWF models [262]. The performance of the integrated oil spill model is assessed through the comparison of model simulations with airborne observations of an oil slick.…”