On the mechanism of interannual variability of the Irminger Water in the Labrador Sea

The mechanism of variability of the North Atlantic subpolar gyre (SPG) and its relation to the North Atlantic Oscillation (NAO) is investigated using an ocean general circulation model. In this study we conducted three model experiments. The first two were forced with idealized positive (NAO(+)) and negative (NAO(-)) NAO-like forcing, including variations at decadal time scales. The third experiment was forced with the National Centers for Environmental Prediction and National Center for Atmospheric Research (NCEP/NCAR) reanalysis. The decadal variability of the volume transport, the sea surface temperature in the North Atlantic Current (SSTA1), and the Irminger Water temperature (IWT) in the NAO experiment have 2-3 times smaller magnitude than in the NAO(+) experiment. The decadal variations in the strength of circulation in the NAO(+) experiment covaries negatively with SSTA1 and IWT anomalies. A similar covariability of these parameters is not found in the NAO(-) simulations. The results from the model experiment forced with the NCEP/NCAR reanalysis from 1958 to 2005 show a shift in the subpolar ocean response to the atmospheric variability in the early 1980s. The amplitude of quasi-decadal variability of the IWT and SSTA1 and their correlation are high after 1980 (r = 0.79) and weaker in the period between 1958 and 1980 (r = 0.1). The IWT after 1980 is well correlated (r = 0.67) to the subpolar gyre transport index (SPGI, defined as the minimum value of the annual mean anomaly of the model barotropic stream function). This correlation is weaker and negative (r = -0.09) in the period from 1958 to 1980. We explain this shift in the covariability of the SSTA1, IWT, and SPGI with the antisymmetric response of the SPG to atmospheric variations at decadal time scale under positive and negative NAO index. The NAO sign changed in the early 1980s from predominantly negative to positive phase. In the 1980s and 1990s the model SPGI variability follows closely the decadal variations of the NAO index with a delay of about 3 years. Similar covariability between the SPGI and the NAO and related negative covariance of SSTA1 and IWT with the SPGI are not observed in the model simulations of the period from 1958 to 1980.