We present definitive observational evidence that the startling change of the Eastern Mediterranean deep circulation observed in winter 1995 and documented by [Roether, W., Manca, B.B., Klein, B., Bregant, D., Georgopoulos, D., Beitzel, V., Kovacevich, V., Luchetta, A., 1996. Recent changes in the Eastern Mediterranean deep water. Science 271, 333–335.] actually started before October 1991. This change involved not only the deep water mass pathways but also the origin and pathways of the water mass spreading in the intermediate layer. We carry out the first unified analysis of the POEMBC-O91 data set, which shows that, differently from the previous decade of the 80s, the Cretan/Aegean Sea was in 1991 the `driving' engine of the intermediate, transitional and deep layer circulations, with Cretan Intermediate Water (CIW), transitional water and Cretan Deep Water (CDW) spreading out from the Cretan Sea into the basin interior. The most important new results are: (a) the Levantine Intermediate Water (LIW) formed inside or at the periphery of the Rhodes gyre is blocked in its traditional westbound route on its density horizons σθ=29.05 and 29.10 kg/m3 by a three-lobe strong anticyclonic structure in the Southern Levantine, which induces a substantial LIW recirculation in the Levantine basin itself; (b) the CIW exiting from the Western Cretan Arc Straits spreads into the Ionian interior on the σθ=29.05–29.10 kg/m3 isopycnal surfaces, thus replacing the LIW confined in the Levantine basin. A branch of CIW flows eastward in the Cretan passage and is entrained by the Ierapetra anticyclone to flow again into the Cretan Sea through the Eastern Cretan Arc Straits; (c) on the horizons σθ=29.15 and 29.18 kg/m3 a transitional water mass of Cretan origin, denser than CIW, and CDW are observed to spread out massively from the Cretan Arc Straits both into the Ionian and Levantine interiors. These isopycnal surfaces rise to much shallower depths in 1991 than in 1987, increasing the salt content of the intermediate, transitional and deep layers. This leads to a massive salt increase in the Ionian below 1200 m, clearly related to lateral advection of the new denser waters of Cretan/Aegean origin, thus contradicting the hypothesis of a vertical salt redistribution proposed by Roether et al.
Read full abstract