Empirical orthogonal function analysis of AVHRR surface temperature patterns: a case study

By means of time series analysis of sea surface temperature (SST) maps it is possible to investigate the temporal and spatial variability of surface currents: such investigation is the basis for the comprehension of the relationship existing between 3D circulation and surface structures. In this work the study of SST variability is focused on the Alboran Sea, Western Mediterranean. The Alboran Sea communicates with the Atlantic Ocean through the Strait of Gibraltar and is bounded on the north by Spain and on the south by Morocco and Algeria. The main oceanographic features of the area are constituted by: 1) the intense baroclinic jet associated to the Atlantic Water stream, forming one or two anti-cyclonic gyres (the westernmost of which is the more persistent and is usually called Western Alboran Gyre, WAG); 2) a strong density and temperature front between Spain and Africa (the Almeria-Oran Front, AOF); 3) mesoscale phenomena associated to the jet and the frontal areas. The MAST-III OMEGA Project (Observing and Modelling of Eddy scale Geostrophic and Ageostrophic circulation) aims at better understanding of this mesoscale system, with particular concern to vertical velocity and its effects on biology. Both remotely sensed data obtained from TM and AVHRR satellite sensors and in-situ measurements were used; the latter were gathered during two cruises: the BIO Hesperides cruise, carried out in the WAG area in October 1996, and the RSS Discovery cruise, carried out in the AOF area in December 1996. The empirical orthogonal function decomposition (EOF) is the method the authors use to analyse SST time series. EOF consists of representing each image of the original sequence on a new basis whose functions are intrinsic to the particular data set because wholly determined by the dataset itself, hence the designation empirical.