Predictability of SST in an Idealized, One-Dimensional, Coupled Atmosphere–Ocean Climate Model with Stochastic Forcing and Advection

Abstract

The predictability of sea surface temperature (SST) is examined through analysis of an idealized, one-dimensional, stochastically forced climate model. The influence on SST predictability of including advection by a constant mean current is investigated. A new mechanism is described whereby predictability is enhanced via a cancellation of stochastically driven noise. For a sufficiently weak advective current the predictability was found to have significant departures from red noise predictability. Bounds on the predictability in the limit of zero advecting velocity were found. The relationship between autocovariance function (or power spectrum in the frequency domain) and predictability is also examined. Perhaps surprisingly, the regions with maximum predictability were not clearly identifiable by their autocovariance function (or power spectrum).