Evaluation of climatological mean surface winds over the Korean Waters simulated by CORDEX regional climate models

Abstract

Surface wind over the ocean not only influence climate change through air-sea interactions but also coastal erosion through changes in wave height and direction. A reliable projection of future changes in surface winds is thus important, and it is is necessary to perform an assessment of their representation in climate models that are widely used for future projections. This study aims to evaluate the climatological mean surface winds over the Korean Waters simulated by fiveCoordinated Regional Climate Downscaling Experiment for East Asia (CORDEX-EA) regional climate models. Compared with the ERA-interim reanalysis data, the CORDEX regional models, except for the HadGEM3-RA model, produce a stronger wind speed over most area of the Korean Waters both in summer and winter. Unlike the other models, HadGEM3-RA, without spectral nudging, underestimates the amplitude of the seasonal mean wind speeds. When examining the wind direction, three out of the five models realistically simulates the observed seasonal reversal of winds (northwesterly in winter and southeasterly in summer) while the other two models show southerly in summer. Our results suggest that large-scale flow information,downscaled to regional models by means of spectral nudging, can significantly affect the surface winds on a regional scaled it is is necessary to perform an assessment of their representation in climate models that are widely used for future projections. This study aims to evaluate the climatological mean surface winds over the Korean Waters simulated by fiveCoordinated Regional Climate Downscaling Experiment for East Asia (CORDEX-EA) regional climate models. Compared with the ERA-interim reanalysis data, the CORDEX regional models, except for the HadGEM3-RA model, produce a stronger wind speed over most area of the Korean Waters both in summer and winter. Unlike the other models, HadGEM3-RA, without spectral nudging, underestimates the amplitude of the seasonal mean wind speeds. When examining the wind direction, three out of the five models realistically simulates the observed seasonal reversal of winds (northwesterly in winter and southeasterly in summer) while the other two models show southerly in summer. Our results suggest that large-scale flow information,downscaled to regional models by means of spectral nudging, can significantly affect the surface winds on a regional scale