Numerical semiprediction of monthly mean sea surface temperature

Abstract

A mathematical model has been developed to specify and predict sea surface temperatures averaged over a long period of time. The model consists of a perturbation type analysis of the thermodynamic equation as applied to the ocean's surface layer. The terms of the equation include the advection of the normal and anomalous isotherms by means of the anomalous and the monthly mean currents, respectively, and the anomalies of heating by evaporation, conduction, and radiation. It is assumed that anomalies of ocean currents can be derived from anomalies of the surface pressure field by making the geostrophic approximation and estimating the drag of the resulting wind on the ocean's surface. For prediction purposes an estimate of the monthly mean sea level pressure for 15 days in advance has been used to arrive at anomalous currents as stated above. Case studies are computed to illustrate the importance of several individual terms in the equation for several seasons. The predictive value of such a model is also illustrated by means of several case studies. The results of these studies suggest that despite the multiplicity of the modeling assumptions such an approach may yield useful monthly mean ocean temperature predictions.