Current data assimilation modelling for oil spill contingency planning

Abstract

A method has been developed to assimilate, in near real-time (15 min updates), current meter data collected by Shell in its operations on the Brunei Shelf Sea. The results from the software are depth mean current flow fields that take into account the requirements of mass and momentum conservation. The data assimilation method minimises, in a least squares sense, a cost function that describes the differences between the measured data and the model counterparts. The minimisation is subject to the weak constraint that the residuals of the shallow water equations are also minimised. Based on current data collected over one year, the data assimilation method was also used to calculate a set of 209 ‘typical’ flow fields representing the most frequent flow conditions in that year. Continuous trajectories of hypothetical spills were computed by integrating particle tracks forward through the year-long current time series along with the corresponding wind data. At each time step, the most appropriate ‘typical’ flow that best represents the measured flow was selected and scaled for use in the integration. The trajectories were used to assess risks of oil landfall.