Decadal deep-water variability in the subtropical Atlantic and convection in the Weddell Sea

Citation
A. Stossel et Sj. Kim, Decadal deep-water variability in the subtropical Atlantic and convection in the Weddell Sea, J GEO RES-O, 106(C10), 2001, pp. 22425-22440
Citations number
70
Language
INGLESE
art.tipo
Article
Categorie Soggetti
Earth Sciences
Journal title
JOURNAL OF GEOPHYSICAL RESEARCH-OCEANS
ISSN journal
2169-9275 → ACNP
Volume
106
Issue
C10
Year of publication
2001
Pages
22425 - 22440
Database
ISI
SICI code
0148-0227(20011015)106:C10<22425:DDVITS>2.0.ZU;2-0
Abstract
This study suggests a link between convection in the Weddell Sea and deepwa ter flows in the (sub)tropical Atlantic. Employing a global sea ice-ocean g eneral circulation model, events of enhanced convection are found to create enhanced outflow of Antarctic Bottom Water (AABW), the associated anomaly of which propagates rapidly along the model's deep western boundary. The pr opagation timescale of the order of a few years to reach the tropical Atlan tic matches baroclinic Kelvin waves with phase speed corrected for the mode l's grid resolution. This feature is consistent with earlier findings on th e baroclinic adjustment to perturbations of the rate of North Atlantic Deep Water (NADW) formation. The mean deep ocean transports, as well as the tim escales and amplitudes of the deep ocean anomalies, are in good agreement w ith observation-based estimates. Upon arrival in the deep western tropical Atlantic, the model AABW anomalies induce strong anomalies in NADW outflow. This behavior suggests that the actual magnitude of NADW outflow across 30 S is highly susceptible to external perturbations on the baroclinic adjust ment timescale. The implied higher susceptibility is supported by recent ob servations that NADW turns mostly eastward between 20 S and 30 S. The NADW anomalies eventually propagate southward and join the Antarctic Circumpolar Current. There is weak indication that the associated anomalies enter the eastern Weddell Gyre to eventually feed back on the critical convection sit e in the Weddell Sea, possibly sustaining the oscillation: However, there i s stronger evidence that the anomalies initially created by the convective events are advected with the model's Weddell Gyre, the advective timescale of which (10-12 years) is consistent with the periodicity of the convective events.