The role of different wind input, whitecap dissipation and quadruplet wave-wave interaction terms in wave evolution in Lake Michigan

In present study, the sensitivity of WAVEWATCH-III model to different wind sources, formulations for wind input, whitecap dissipation and quadruplet wave-wave interaction was investigated during both fair weather and extreme condition of Hurricane Sandy in Lake Michigan. Wind data sets with finer spatial resolution (GLERL and GEM) led to better agreement with in-situ data, while ERA5 with coarser spatial resolution underestimated wind speed both in deep and shallow waters. Preparatory evaluation of model skill in reproducing wave bulk parameters indicated that the highest error was obtained when ST4 package and ERA5 wind data were employed in model. The calibration procedure was conducted by employing a combined error index. Simulated wave bulk parameters with optimum tuning values for GLERL and GEM wind data were in well agreement with in-situ data; however, GEM outperformed GLERL during extreme values of Hs and Tm in deep water buoys. Moreover, GLERL wind data slightly improved model skill in Hs and Tp estimation in shallow water buoys. Although more advanced methods were considered for quadruplet wave-wave interaction, DIA could be sufficiently reliable method for hindcasting wave bulk parameters. The DIA method systematically overestimated wave energy at low frequencies which was alleviated when GMD or GQM method was employed. Obtained directional wave spectra by using ST1 and ST6 packages combined with DIA were directionally broad as well as less energetic in peak frequency; however, GMD, GQM and WRT methods were led to more concentrated high energy content. In general, the quality and quantity of modeled wave spectra were in a reasonable agreement with observed one after model calibration. •ERA5 underestimates wind speed at buoys, while GLERL and GEM are in more agreement with in-situ data in Lake Michigan•High error indices for Hs, Tp and Tm hindcast were obtained when ST4 package and ERA5 were applied in WWIII model.•ST6 was calibrated against observed Hs and Tm at deep water buoys during fair weather as well as extreme Hurricane Sandy.•GLERL outperformed GEM in Hs simulation at shallow buoys; but GEM better estimated Hs and Tm at deep buoys.•DIA is a reliable method in wave bulk parameters predictionSnl