Validation of Hydrodynamic Ocean Models Using Empirical Data For The Purpose of Larval Dispersal Modelling

Larval dispersal modelling and many other applications of hydrodynamic models require extensive time series of accurate high-resolution simulations of the ocean state, particularly current velocities and seawater temperature.... [ view full abstract ]

Larval dispersal modelling and many other applications of hydrodynamic models require extensive time series of accurate high-resolution simulations of the ocean state, particularly current velocities and seawater temperature. Ocean current velocities drive the dispersal of planktonic organisms and other passive drifters, while seawater temperature is a key parameter in the survival of propagules. In this study we propose a method to assess the accuracy of hydrodynamic ocean models by using empirical data. The ocean models we tested are HYCOM (Hybrid Coordinate Ocean Model) developed by the Centre for Ocean-Atmospheric Prediction Studies (COAPS), and BRAN (Bluelink ReANalysis) and RIBBON models developed by Commonwealth Scientific and Industrial Research Organisation (CSIRO). BRAN is a global model that runs on a z-level grid, with coarse coastal horizontal resolution of 0.1°. HYCOM is a global ocean model that uses a hybrid grid for improved coastal simulations and has a 0.08° horizontal resolution. RIBBON is a regional model on a curvilinear σ-level grid, covering the entire Australian continental shelf at a horizontal resolution between 0.01° and 0.1°. We validated the predictions of these hydrodynamic models for seawater temperature, u and v components of current velocity in the ocean coastal domain of Australia against in situ measurements from the Australian Network of National Moorings (ANMN). This method provides a spatially explicit test of model accuracy at single points in space.