AUTHOR=Gao Yang , Schmitt François G. , Hu Jianyu , Huang Yongxiang TITLE=Probability-based wind-wave relation JOURNAL=Frontiers in Marine Science VOLUME=9 YEAR=2023 URL=https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2022.1085340 DOI=10.3389/fmars.2022.1085340 ISSN=2296-7745 ABSTRACT=

In a fully developed sea, the significant wave height (Hs) and wind speed (U10) are conventionally related to a pure quadratic equation. This relation is often violated, since in the field the measured local Hs is often contaminated by the swell, which is propagated from distant places. Therefore, a swell partition is required before the establishment of the wind-wave relation. The Spectra Energy Partition (SEP) is regarded as the best way to isolate the swell and the wind wave relation: it is based on the identification of a separation frequency in the ocean wave spectrum. However, for most field observations, the wave spectra information is unavailable, and thus the SEP is inapplicable. This work proposes a probability-based algorithm to identify the averaged swell without knowing the wave spectrum a priori. The local wind-wave relation is established by either a linear or an energy-conserved decomposition. We also find that the local wind-wave relation is a power-law when the wind speed U10 is above 4 m/s. The proposed method is first validated by applying the SEP method to buoy collected wave spectra data. The global pattern of the swell and the local wind waves is retrieved by applying the proposed method to a 17-year wind and wave database from the JASON satellite. Strong seasonal and spatial variations are obtained. Finally, a prediction model based on the empirical wind-wave relation is shown to successfully retrieve the wave field when the wind field is available.