Operational Oil Spill Modelling: From Science to Engineering Applications in the Presence of Uncertainty - Mathematical Modelling and Numerical Simulation of Oil Pollution Problems - page 120

5.10 Conclusions
This chapter has presented a discussion of the major challenges and opportunities
involved in creating oil spill forecasting systems that account for uncertainty. Forecast
uncertainty is added with each modelling step and parameterization, from the initial
wind forecast to the oil spill diffusivity. There is a need for emergency responders
to have a real-time understanding of the uncertainty of the forecast for today's spill,
which might not be well predicted by hindcast studies at different times or locations.
A systematic approach to including uncertainty at all levels in real-time forecasts has
been proposed (Sect. 5.6 ), but there are significant engineering challenges to putting
these ideas into operation at any particular location. Probably the most daunting
challenge for many operational agencies will be bureaucratic rather than scientific or
engineering: for a viable multi-model uncertainty forecast system, a research team
needs direct access to an operational coastal hydrodynamic model to run multiple
simulations with different wind forecasts. However, if the uncertainty associated
with wind, wave, and hydrodynamic forecasts can be parameterized into a range of
diffusivities for an oil spill model, then an approximation of the multi-model uncer-
tainty can be directly integrated into an ensemble approach to the oil spill modelling
(Sect. 5.9 ). The key point is that uncertainty in wind forecasts and the surface water's
response to the wind are the critical drivers of uncertainty [ 27 , 48 ], and therefore
must be included in any uncertainty evaluation system—either directly through a
multi-model approach, or indirectly through hindcast analyses and parameterization
of particle diffusivity for a forecast oil spill model.
Acknowledgments The work of X. Hou and B.R. Hodges is based upon work supported by the
Research and Development program of the Texas General Land Office Oil Spill Prevention and
Response Division under Grant No. 13-439-000-7898 and in part by a grant from BP/The Gulf of
Mexico Research Initiative. A. Orfila and J.M. Sayol would like to thank the support fromMICINN
through Project CGL2011-22964. J.M. Sayol is supported by the PhDCSIC-JAE program cofunded
by the European Social Fund (ESF) and CSIC.
References
1. Abascal, A.J., Castanedo, S., Medina, R., Losada, I.J., Alvarez-Fanjul, E.: Application of HF
radar currents to oil spill modelling radar. Mar. Pollut. Bull. 58 , 238-248 (2009)
2. Abascal, A.J., Castanedo, S., Medina, R., Liste, M.: Analysis of reliability of a statistical oil
spill response model. Mar. Pollut. Bull. 60 , 2099-2110 (2010)
3. ASCE.: State-of-the-art review of modelling transport and fate oil spills, ASCE committee
on modelling oil spills water resources engineering division. J. Hydraul. Eng. 122 , 594-609
(1996)
4. Beegle-Krause, C.J.: GNOME: NOAA's next-generation spill trajectory model. In: Oceans'99
MTS/IEEE Proceedings Conference Committee, vol. 3, pp. 1262-1266 (1999)
5. Bennett, J.R., Clites, A.H.: Accuracy of trajectory calculation in a finite-difference circulation
model. J. Comput. Phys. 68 , 272-282 (1987)
6. Berry, A., Dabrowski, T., Lyons, K.: The oil spill model OILTRANS and its application to the
Celtic Sea. Mar. Pollut. Bull. 64 , 2489-2501 (1990)