State-of-the-art techniques in oil spill modeling

Abstract

The status and possible research directions of oil spill modeling are presented here. The physical and chemical processes that take place in oil spills are explained as is their role in the design of an ideal oil spill model. An ideal oil spill model for forecasting must support rapid response, contingency planning, and training. Accurate, full-dimension, real-time prediction of hydrodynamic calculations and oil movements require intensive computations and computing power. Presently the most promising computational platform appears to be a vector supercomputer that has been given essential geographical, bathymetric, and tidal data before the spill. Parallel processing machines could be used in the same way; hut present codes, written for sequential machines, must he changed substantially to take advantage of the parallel architecture. Super minicomputer technology is advancing rapidly and should soon be able to run the numerically intensive hydrodynamic codes, bringing the advantages of portability and low cost. Since the present personal computer oil spill models are not able to make accurate hydrodynamic current predictions, they cannot match the accuracy of more computationally intensive models.