| description abstract | 2-yr-long daily gridded field of sea surface temperature (SST) in the Atlantic centered for the year 2013 is projected onto orthogonal components: its mean, six harmonics of the year cycle, the slow-varying contribution, and the fast-varying contribution. The periodic function defined by the year harmonics, referred to here as the seasonal harmonic, contains most of the year variability in 2013. The seasonal harmonic is examined in its spatial and temporal distribution by describing the amplitude and phase of its maxima and minima, and other associated parameters. In the seasonal harmonic, the ratio of the duration of warming period to cooling period ranges from 0.2 to 2.0. There are also differences in the spatial patterns and dominance of the year harmonics?in general associated with regions with different insolation, oceanic, and atmospheric regimes. Empirical orthogonal functions (EOFs) of the seasonal harmonic allow for a succinct description of the seasonal evolution for the Atlantic and its subdomains. The decomposition can be applied to model output, allowing for a more incisive model validation and data assimilation. The decorrelation time scale of the rapidly varying signal is found to be nearly independent of the time of the year once four or more harmonics are used. The decomposition algorithm, here implemented for a single year cycle, can be applied to obtain a multiyear average of the seasonal harmonic. | |
