MEDSLIK oil spill model predicts the transport, fate and weathering of the oil spill. It uses the atmospheric, wave and hydrodynamical data from predictive models in combination with information on the
- oil spill location
- date and time of the oil spilled at sea
- duration of integration
- volume or rate of oil spilled at sea
- nature and characteristics of the oil spilled at sea, using either the REMPEC data base with 240 different oil type characteristics or the API number
taking into account the
- 3D currents
- Stoke’s drift by waves
- wind drift factor and wind drift angle
- horizontal diffusivity that may be estimated using the currents will be provided by the hydrodynamical predictive models
- vertical diffusivity
- depth of upper mixed layer
- vertical density of the sea water in case of oil spill below the sea surface
The MEDSLIK oil spill model incorporates:
- the evaporation,
- viscosity changes and
- dispersion in water column adhesion to coast and sedimentation.
The oil spill movement is simulated using a Monte Carlo method. The pollutant is divided into a large number of Lagrangian parcels of equal size. At each time step, each parcel is given a convective and a diffusive displacement.
Τhe oil is considered to be consisted by a light evaporative component and a heavy non evaporative component. Emulsification is also simulated, and the viscosity changes of the oil are computed according to the amount of emulsification and evaporation of the oil. The model simulates slick transports taking into account that the movement of the surface slick is governed by currents, waves, Stoke’s drift and wind, while the diffusion of the slick is simulated by a random walk (Monte Carlo) model. The oil may be dispersed into the water column by wave action, while the dispersed oil is moved by currents only. Mechanical spreading of the initial slick is also included.The number of parcels that is used by the MEDSLIK model to form the oil spill may ranges from 10.000 up to 500.000, while the water column structure is described by 9, adjustable by the user, vertical layers to the relevant hydrodynamic/oceanographic data provided by the predictive models.
Moreover, the MEDSLIK oil spill model prediction length may vary from few hours up to 3 weeks, but using the ‘restart” facility of the model, the oil spill prediction length can be extended further depending on the end user application requirements and the forcing availability.
MEDSLIK oil spill model has been implemented in operational mode in the Mediterranean Sea, while a pre-operational mode of MEDSLIK is already used in the Baltic Sea and the Black Sea. At present, the source of atmospheric forcing is SKIRON high frequency data, as well as the low resolution ECMWF, while the oceanographic forcing is taken from the Copernicus marine service and the downscaled CYCOFOS regional ocean forecasting systems. The required wave data are received from CYCOFOS wave model-WAM for the Mediterranean and the Black Sea.
MEDSLIK has been successfully used for risk assessment from the offshore planning of the hydrocarbon exploration in the Eastern Mediterranean Levantine basin, where oil spill hindcast simulations were performed for 4 years at 10 locations proposed by the Italian ENI.
Finally, MEDSLIK oil spill model has been extensively validated in the Mediterranean Sea, both in real oil spill incidents (i.e. during the Lebanese oil pollution crisis in summer 2006, the biggest oil pollution event in the Eastern Mediterranean so far) and through EMSA warning reports and satellite images (ASAR)