Modeling double-peaked spectra by combining two dimensionless JONSWAP models

Abstract

Modeling a double-peaked spectrum naturally derives the significant wave heights Hs1 for the swell system and Hs2 for the wind-sea system. However, few studies have examined whether Hs1 and Hs2 are consistent with the observed significant wave heights of swell Hss and wind-sea Hsw. The purpose of this research is to develop a high-precision double-peaked spectral model with the spectrum-related significant wave heights Hs1 and Hs2 that closely match the observed Hss and Hsw calculated by the wind-sea and swell separation method, respectively. A novel double-peaked spectral model based on the combination of two dimensionless JONSWAP spectra is found to be advantageous over the classical GH98 model (Guedes Soares and Henriques, 1998) and the APA19 model (Akbari et al., 2019) in the applications to three field datasets of coastal China. Specifically, the present model is comparable to the APA19 model in terms of fitting accuracy but with less computation, and both are substantially superior to the GH98 model. The present model outperforms the APA19 and GH98 models regarding the correlation between spectrum-related and observed significant wave heights. Furthermore, the ordinates of the peaks of the present model are precisely consistent with those of the observed spectrum, whereas they are frequently underestimated or overestimated by the GH98 model and the APA19 model.