Leading decadal variability and climate linkages in a gridded dataset of North Atlantic extratropical surface temperature, salinity, and density from 1896–2018

Andrew Ronald Friedman1, Leon Chafik2, Gilles P Reverdin3, N. Penny Holliday4 and Gabriele C Hegerl1, (1)University of Edinburgh, Edinburgh, EH9, United Kingdom, (2)Stockholm University, Department of Meteorology, Stockholm, Sweden, (3)Sorbonne Université - CNRS/IRD/MNHN, LOCEAN, Paris, France, (4)National Oceanography Centre, Southampton, United Kingdom
Abstract:
We examine the decadal variability in a new binned product of North Atlantic sea surface salinity (SSS) and temperature (SST) from 1896–2018 between 40°N and 70°N. The leading principal components (PCs) of detrended normalized SSS and SST show a monopole structure, most coherent in the central part of the subpolar gyre. The leading PC (PC1) time series have large variance at decadal and multidecadal timescales and are highly covarying, with slightly stronger autocorrelation in SSS. They have maxima around 1940 and the early 2000s, and minima around 1970s. Lag correlation analysis of the leading SSS PC1 shows a progression to the northeast. Using composite analysis, we investigate the evolution of the PC1 time series during its transition decades from negative to positive phase in the 1920s, positive to negative phase in the 1960s, and negative to positive phase in the 1990s. The roles of surface forcing and linkages with large-scale atmospheric and ocean modes of climate variability are assessed. We also examine the relationships to sea ice concentration in different SPG sub-basins. PC1 is significantly negatively correlated with Davis Strait / Labrador Sea ice concentration with a lag of 1-2 years.