NOWCASTING INFRAGRAVITY WAVE HEIGHT WITHIN A HARBOUR USING AN ARTIFICIAL NEURAL NETWORK

Abstract

Infragravity (IG) waves, sometimes known as “long waves” “low frequency waves” or “subharmonic gravity waves” are waves that are forced in deeper water through subharmonic interactions of sea and swell wave components. These waves are typically quite small (a few centimetres to decimetres) but can excite harbour seiching and amplification due to resonance. This can cause vessels to surge and sway at their berth, straining mooring lines and making loading and unloading difficult. This paper describes the design, training and application of a Bayesian regularized Artificial Neural Network (ANN) to accurately predict (“nowcast”) IG wave height outside and inside Port Kembla from deep-water swell wave spectral characteristics measured at Port Kembla Waverider Buoy (PKWRB), NSW coast (Williams, 2019).