This dataset represents a high-resolution (1/12° in the horizontal) regional ocean simulation of the North Atlantic, designed to explore the impact of atmospheric winds magnitude and variability on subsynoptic (<2 days) and synoptic (2-10 days) timescales on the ocean circulation in the region. The MITgcm (Massachusetts Institute of Technology General Circulation Model) ocean general circulation model was used, forced by repeated year forcing conditions from JRA55-do, corresponding to 1 May 2003 to 1 May 2004, and lateral boundary conditions from Arctic STate Estimate (ASTE) dataset (Nguyen et al., 2021), all these experiments lasted for 15 years. In total, this dataset contains three sensitivity experiments: (1) WIND_LF_SYNOP - where the variability of surface forcing winds filtered out on periods shorter than 2 days; (2) WIND_LF - where the synoptic plus higher frequency variability of surface forcing winds were filtered out; and (3) WIND_LF_SCALED - experiment where the synoptic and higher frequency variability were filtered out; however, the magnitude of the wind speed was scaled so the total wind energy input was the same as in the control run forced by reanalysis winds. Full 3D U/V/T/S fields and 2D fields containing information about mixed layer depth, sea ice, sea surface height, heat and freshwater surface fluxes are provided every 5 days. The outputs were generated by the University of Oxford under the Natural Environment Research Council (NERC) project SNAP-DRAGON: Subpolar North Atlantic Processes - Dynamics and pRedictability of vAriability in Gyre and OverturNing (grant reference NE/T013494/1).