The south coast of Korea is vulnerable to coastal disasters, such as storm surges, high waves, wave overtopping, and coastal flooding caused by typhoons. It is imperative to predict such disastrous events accurately in advance, which requires accurate meteorological forcing for coastal ocean modeling. In this study, to acquire accurate meteorological data as important forcing variables for the prediction of storm surges and waves, we examined the forecast performance and applicability of a next-generation global weather/climate prediction model, the Model for Prediction Across Scales (MPAS). We compared the modeled surface pressure and wind with observations on the south coast of Korea for three typhoons that damaged Korea in 2020—Bavi, Maysak, and Haishen—and investigated the accuracy of these observations with the MPAS prediction. Those meteorological forcing variables were then used in the tightly coupled tide-surge-wave model, Advanced CIRCulation (ADCIRC) and the Simulating Waves Nearshore (SWAN) for the simulation of a typhoon-induced storm surge and wave. We also performed the hindcast of the wave and storm surges using a parametric tropical cyclone model, the best-track-based Generalized Asymmetric Holland Model (GAHM) embedded in ADCIRC+SWAN, to better understand the forecast performance and applicability of MPAS. We compared the forecast results of the typhoon-induced wave and storm surges with their hindcast in terms of the time-series and statistical indices for both significant wave height and storm surge height and found that wave and storm surge prediction forced by MPAS forecast provides a comparable accuracy with the hindcast. Comparable results of MPAS forcing with that of hindcast using best track information are encouraging because ADCIRC+SWAN forced by MPAS forecast with an at most four-day lead time still provides a reasonable prediction of wave and storm surges.