Abstract
Based on available data, a 100-year series of annual sea surface temperature (SST) is established for a given area in the North Atlantic. The oscillations of the annual SST anomalies are characterized as small oscillations about a stable state of equilibrium obviously determined by the heat balance.
The behavior of such a system can be described by linear integro-differential equations with constant coefficients. It is shown that the power spectrum of a second-order autoregressive linear stochastic process reveals the main features of the spectrum of the given SST time series. The observed low frequency significant peak of the annual SST anomaly spectrum is assumed to be the oceanic integral response to the white noise atmospheric input on shorter time scales. The low frequency location of the peak, corresponding to a period of 25–30 years, is physically explained by the large thermal inertia caused by the large heat capacity of the ocean.
A thermal feedback of low frequency SST oscillations to the atmosphere explains the low frequency latitudinal oscillations of the annual mean tropospheric thermal field, including the polar frontal zone from the North Atlantic toward Europe.
