AbstractAdded value (AV) from dynamical downscaling has long been a crucial and debatable issue in regional climate studies. To assess AV generated by COSMO Climate Local Model (CCLM), a model-reconstructed hindcast with 7 km grid resolution was compared against the forcing data set ERA-Interim (ERA-I) over the Bohai Sea and the Yellow Sea, with satellite and in situ observation as reference. Both quantitative metrics and qualitative assessments have been used in the investigation of AV by CCLM. Land surface winds, extreme winds, a typhoon, a cold surge and a vortex street have been selected in the assessment process. Statistical analysis on surface winds reveals that high-resolution CCLM hindcast can add value to ERA-I in reproducing wind intensities and direction, probability distribution and extreme winds mainly at mountain areas, which may be related to the highly resolved and accurate description of terrain roughness length and orographic barriers in CCLM. With respect to atmospheric processes, CCLM outperforms ERA-I in resolving detailed temporal and spatial structures for phenomena of a typhoon and of a cold surge; CCLM generates some orography-related phenomena such as a vortex street, which is not captured by ERA-I. These AVs demonstrate the utility of the 7-km resolution CCLM for regional and local climate studies and applications.