124
H.E.M. Meier and A. Höglund
particular information without saving the complete fields. This illustrates the need
to plan the simulation and implement the right analysis already in the model.
4.4.2 Multi-tracer Approach
Until now we focused the discussion on the usage of individual tracers. However,
it is possible to use also several passive tracers representing different states of any
substance. The idea is to connect sinks and sources of different tracers. The transformation between the tracers represents a physical, chemical or biological process,
and its calculation might be combined even with a third tracer. For instance, one
tracer is used for the description of the oil slick at the surface and another one for oil
vertically mixed within the water column. The transformation between the oil slick
and other oil compounds would represent vertical mixing due to surface waves. This
process will depend on the wave height. In the other direction buoyant oil will be
floated up to the surface.
This approach is realized within the Multiphase Oil Spill Model (MOSM)
(Tkalich et al. 2003). The MOSM model uses a regular grid and six fields: one
two-dimensional tracer for the oil slick thickness at the surface, three 3D tracers
for emulsified oil, particulate oil and dissolved oil in the water column, and finally
two 2D fields for particulate oil and dissolved oil buried in the bottom sediments
which are not advected. The advection of the surface tracer is not only done with
the surface currents but also with a fraction of the wind. A number of processes
can transform the oil compounds into each other. Oil can also leave the system by
evaporation. The composition of the oil, together with many other parameters, determines coefficients for the rate of transfers and evaporation. For more details of
oil spill modelling the reader is referred to Chap. 11 of this book.
Tkalich et al. (2003) motivated the usage of Eulerian tracers with “Nowadays,
an oil slick dynamics model can afford to routinely use such accurate and physically relevant formulation as the Navier–Stokes equations.” Another motivation is
to make it easier to couple the oil spill model to other models. For instance, the
MOSM model was coupled with a food chain model to analyse the effects of an
oil spill on the food web (Gin et al. 2001). The food chain model contains phytoplankton, zooplankton, small fish, large fish and benthic invertebrates. Although the
coupling of the models would not have been impossible if the oil spill model was
Lagrangian, it would have been more complicated.
Finally the very interesting question arises whether the model performed better
than other more traditional models. Although Tkalich et al. (2003) presented some
validations, unfortunately no details were given. It was concluded that “test simulations of oil slick dynamics, as well as transport and kinetics of the oil phases in
the water column, show good consistency with empirical data and other models”
(Tkalich et al. 2003).
H.E.M. Meier and A. Höglund
particular information without saving the complete fields. This illustrates the need
to plan the simulation and implement the right analysis already in the model.
4.4.2 Multi-tracer Approach
Until now we focused the discussion on the usage of individual tracers. However,
it is possible to use also several passive tracers representing different states of any
substance. The idea is to connect sinks and sources of different tracers. The transformation between the tracers represents a physical, chemical or biological process,
and its calculation might be combined even with a third tracer. For instance, one
tracer is used for the description of the oil slick at the surface and another one for oil
vertically mixed within the water column. The transformation between the oil slick
and other oil compounds would represent vertical mixing due to surface waves. This
process will depend on the wave height. In the other direction buoyant oil will be
floated up to the surface.
This approach is realized within the Multiphase Oil Spill Model (MOSM)
(Tkalich et al. 2003). The MOSM model uses a regular grid and six fields: one
two-dimensional tracer for the oil slick thickness at the surface, three 3D tracers
for emulsified oil, particulate oil and dissolved oil in the water column, and finally
two 2D fields for particulate oil and dissolved oil buried in the bottom sediments
which are not advected. The advection of the surface tracer is not only done with
the surface currents but also with a fraction of the wind. A number of processes
can transform the oil compounds into each other. Oil can also leave the system by
evaporation. The composition of the oil, together with many other parameters, determines coefficients for the rate of transfers and evaporation. For more details of
oil spill modelling the reader is referred to Chap. 11 of this book.
Tkalich et al. (2003) motivated the usage of Eulerian tracers with “Nowadays,
an oil slick dynamics model can afford to routinely use such accurate and physically relevant formulation as the Navier–Stokes equations.” Another motivation is
to make it easier to couple the oil spill model to other models. For instance, the
MOSM model was coupled with a food chain model to analyse the effects of an
oil spill on the food web (Gin et al. 2001). The food chain model contains phytoplankton, zooplankton, small fish, large fish and benthic invertebrates. Although the
coupling of the models would not have been impossible if the oil spill model was
Lagrangian, it would have been more complicated.
Finally the very interesting question arises whether the model performed better
than other more traditional models. Although Tkalich et al. (2003) presented some
validations, unfortunately no details were given. It was concluded that “test simulations of oil slick dynamics, as well as transport and kinetics of the oil phases in
the water column, show good consistency with empirical data and other models”
(Tkalich et al. 2003).