Author(s): Connor Jordan; Athanasios Angeloudis

Linked Author(s):

Keywords: Shallow water equations; Tidal resource assessment; Parameter estimation; Bottom friction; Manning coefficient

Abstract: Assessments of the potential of tidal stream energy at a site often rely on only one or two Acoustic Doppler current profilers (ADCPs) for validation. In numerical modelling, an optimal uniform value for the friction field is typically determined by sensitivity analysis. However, at numerous sites, the resource is subject to significant spatio-temporal variation. We investigate the impact of limited validation data and the assumption of uniform friction for numerical model calibration with a case study at the world’s current highest capacity commercial array deployment. The best performing value of a uniform Manning coefficient is determined using a sensitivity study type approach, and then compared to an adjoint-based calibration of the model where the Manning coefficient can spatially vary. The model is progressively updated to emulate the process of obtaining more ADCP data over several surveys as funding is made available. We find that a uniform Manning coefficient cannot simultaneously provide good agreement over multiple ADCP datasets collected at the site. In adjoint-based calibration, when relying on initial data, it is susceptible to overfitting that is not corrected for until later datasets are incorporated. It is therefore critical to obtain multiple ADCP datasets to reduce uncertainty, improve model accuracy and in turn, improve array design at an early stage.

DOI: https://doi.org/10.64697/978-90-835589-7-4_41WC-P1740-cd

Year: 2025