Die Kuste, 81 (2014), 255-271
262
In case of oil — depending on the availability of information on die kind of oil —
SeatrackWeb gives the possibility to choose just an oil class (light, medium, heavy) or a
specific oil (e.g. marine diesel, IFO 450, Bunker B, etc.). This choice has for example con
sequences on the rate of evaporation and emulsification.
Although die map is only 2-dimensional die drift simulation is 3-dimensional, mean
ing that it is possible to define an outiet in a certain depth and that the substance is dis
persed in die water column if not prevented by buoyancy. The deptii of a particle is color
coded according to a legend shown in die lower right corner of the main window.
Furtiier, die user has several options for analyzing die results. For example it is
possible: to zoom in and out, add layers like for example traffic separation schemes, go
forward and backward in time, show an animation, plot die trajectory of all particles or
only of die barycenter of the particles, show wind and current data, and save map images.
By saving die case die simulation result may be shared with otiier SeatrackWeb users or
loaded in otiier systems tiirough suitable interfaces. It is also possible to save tables and
graphs showing die amount/percentage of oil at die surface, stranded, dispersed,
emulsified, etc.
If the pollution source is unknown and a potential originator is searched for, AIS ship
position can be loaded and displayed in combination with die drift simulation results.
This helps to preselect ships to inspect. Displaying die EMSA provided oil spill detec
tions in satellite images is also possible and helps finding possible polluters by backward
simulations.
3 Results
In this section the performance of SeatrackWeb is demonstrated based on some real cas
es. First die results of an oil spill caused by the average of the cargo vessel “Full city” in
the Skagerrak area in 2009 are shown. Then results of drifting objects, namely containers
in the German Bight in 2012, are presented.
3.1 Ship average in Skagerrak area
On 30 July 2009 the cargo-vessel “Full City” anchored near die Norwegian coast in the
Skagerrak area. During strong gale winds the anchor flukes broke off and die ship started
to drift towards Sastein Island, where it ran aground by night losing about 300 tons of
IF180 bunker oil (BROSTROM et al. 2011). Oil response action started next morning, but
could not prevent a widespread pollution of the Norwegian coast causing ecological and
economic damage. Drift models were used to predict die oil trajectory and a comparison
of three model results — OD3D, SeatrackWeb and BSHdmodL - was published by
BROSTROM et al. (2011) afterwards. All models showed good agreement witii observa
tions. DAMSA used SeatrackWeb witii HIROMB ocean model data and HIRLAM wind
data. In this section die “Full City” case is considered again using SeatrackWeb with
BSHcmod ocean model and GME+LME wind forcing data. The same forcing data was
used by die drift model BSHdmodL in the comparison in BROSTROM et al. (2011). Thus,
differences in die oil spill trajectory are only due to the drift model.
In this paper SeatrackWeb uses die same initial setup as described in BROSTROM et al.
(2011) including some additional uncertainty spreading due to wind as proposed in the