The model experiments are carried out for the global ocean for estimation of the volume transport and its nature as a part of global conveyer-belt circulation. It was shown that the mid-depth and deep layer currents exhibit thermohaline forced and wind stress-modified variation that propagates between southern hemisphere and northern hemisphere. Since the surface heat flux forcing must follw planetary scale dynamics, the disturbances must propagate along deep layer western boundary, the disturbances originated in the southern hemisphere is confined in the western part of the basin, i.e., the thermohaline forced cuurent from southern hemisphere cross the equater near New Guinea and the disturbances originated in the northern hemisphere propagates from the western boundary of the northern Pacifc through the Indonasian seas into the Indian Ocean.
The Indonesain throughflow consists of not only the wind-driven surface flow, but also the thermohaline driven and wind-stress-modified deep layer flow that changes its strength and direction in seasons. The volume transport of the vertically integrated volume transport in the Indonesian Throughflow showed its maximum of 23 Sv is in February and its minimum transport of 11 Sv is in June due to seasonally changing deep layer flow. Although the estimated valus is consistent with the notion of global conveyer belt circulation that is forced by sinking of water in the north Atlantic Ocean, this estimate should be further examined by finer resolution model because the northward flow shown in this coase model resolution may be reflected at the bottom sill and reduced in finner resolution global ocean circulation model.
Acknowledgements
The authors wish to acknowledge Mr. Shinichi Ishii, executive director of Japan Marine Science and Technology Center, who has encouraged the authors to conduct Jamstec expeditions in the early stage of JAMSTEC-WOCE cruise in the western tropical Pacific Ocean from 1990 to 1992. The authors have gained many insights from the ocean observations that has became the prime motivation of this numerical simulation. Several colleagues, including Kazuhiro Ban and Hiroshi Ishikawa, have patiently listened to our expositions of this work and have offered several helpful suggestions and many computational efforts using CRAY supercomputer in Mineapolis, Minesota and NEC SX4 supercomputer in JAMSTEC, Yokosuka. We are grateful to Mr Kiyoshi Otsuka of JMASTEC Computer Center who has provided us the generous CPU times in order to port OPYC code from CRAY to NEC SX4. This research was supported by the tropical ocean observation program under Science and Technology Agency of the Japanese government.
