DOI: 10.1029/2026jc024701 ISSN: 2169-9275

Mode‐2 Nonlinear Internal Wave Generation and Propagation on an Energetic Continental Shelf

Yankun Gong, Matthew D. Rayson, Nicole L. Jones, Gregory N. Ivey

Abstract

While mode‐2 nonlinear internal waves (NLIWs) can play an important role in mixing and transporting material, their generation and evolution remain less well‐understood than their mode‐1 counterparts. Field observations from a mooring in 330 m water on the Australian North West Shelf revealed the frequent occurrence of mode‐2 NLIWs, with leading wave amplitudes of 23 8 m, followed by a train of trailing waves, and a propagation direction of 54° 12° SE (i.e., inshore). While these mode‐2 NLIWs typically occurred every 12–12.4 hr, they were not directly linked to the local tidal forcing. Furthermore, the amplitude of the leading waves increased when the background stratification shifted from a 2‐layer to a 3‐layer regime (a structure influenced by the passage of TC Veronica ). We used a non‐hydrostatic (NH) ocean model to understand the mode‐2 generation and propagation process and its link to the background density stratification and forcing. We first ran a series of 2‐dimensional (2D) NH model runs, driven by both remote internal tides (RIT) with a low‐frequency phase modulation and local barotropic forcing. The 2D modeling demonstrated that non‐phase‐locked mode‐1 remotely‐generated internal tides (RITs) scattered energy to high mode beams at the shelf break. The shoreward‐propagating internal wave beams then interacted resonantly with the double‐pycnocline (3‐layer) stratification to ultimately generate mode‐2 NLIWs in shallower water at depths of around 500 m. Using knowledge from these 2D simulations, we then ran a 3D NH model with phase‐modulated RITs and local barotropic forcing, and the model results were in good agreement with the observations. The crest length of the mode‐2 NLIWs was ∼5 km and the crest width was less than 1 km in the shelf waters, as determined by the correlation length scales calculated from the numerical model results.