Evaluation of parametric tropical cyclone wind models for representing storm surge and wave heights in the Queensland region

Accurate representation of tropical cyclone (TC) surface wind and atmospheric pressure fields is crucial for numerical modelling of storm surges and extreme ocean (wind-)waves to adequately predict and inform related impacts to coastal communities, infrastructure, and ecosystems, yet it remains a challenge. This study evaluates three parametric TC wind models (the Kepert, McConochie, and Chavas-Lin-Emanuel (CLE15) models) alongside surface wind fields from two dynamical regional atmospheric reanalyses (one convection permitting). While most comparable studies primarily evaluate surface wind models against point wind observations, we instead assess how model choice and parameterisation propagate through coupled surge-wave simulations to influence the prediction of storm surges and wind-waves. To do so, we leverage an unstructured mesh coupled hydrodynamic-wave model previously developed to support improved coastal hazard characterisation for Australia. We dynamically simulate storm surges, tide and wind-wave fields for eight historical TCs in the Queensland/Great Barrier Reef region and validate the results against storm surge (tide gauge) and storm wave (buoy) observations. The findings highlight how storm surge and especially wind-wave fields allow for a more integrated, albeit indirect, evaluation of the wind fields compared to direct in-site wind measurements. The findings also illustrate the utility of parametric TC wind models, since even relatively high-resolution, convection permitting regional dynamical reanalyses are limited in their ability to sufficiently resolve the inner-core winds (particularly of intense TCs) for coastal hazard applications such as predicting coastal extreme sea levels. All three parametric models markedly improve the peak surge and wave estimates compared to the regional atmospheric reanalyses. However, there are notable differences among them. The McConochie (double vortex) model exhibits the largest biases, producing an overly broad wind field that overestimates the spatial extent, timing, and duration of storm surge and wave events. The CLE15 and Kepert models demonstrate comparable performance and strong suitability for TC-induced surge and wave modelling in Queensland, with CLE15 model yielding appreciably more accurate and consistent predictions across the cases examined. These results provide guidance for model selection in TC hazard assessments, indicating that physics-based radial wind profiles (like CLE15) can offer improved predictive skill over regionally derived empirical models and reanalysis wind fields.