Turbulence modelling using 3DVAR data assimilation in laboratory conditions

The viability of turbulence parameter estimation in a numerical model using 3DVAR data assimilation technique is explored in this research. Water currents measured in a physical model are assimilated into the numerical model DIVAST in order to improve prediction skill of the model in regions where turbulent processes are of importance. The performance of two turbulence closure schemes, the standard k-e model and the Prandtl mixing length model, is investigated. The assimilation of the model-predicted velocity and laboratory observations significantly improves model predictions for both turbulence schemes. The research further demonstrates how 3DVAR can be utilized to identify and quantify shortcomings of the numerical model and consequently to improve forecasting by correct parameterization of the turbulence models. Such improvements may greatly benefit physical oceanography in terms of understanding and monitoring of coastal systems and the engineering sector through applications in coastal structure design, marine renewable energy and pollutant transport.