A model study of temperature anomaly propagation from
the subtropics to tropics within the South Atlantic thermocline
Alban Lazar
Laboratory for Hydrospheric Processes, USRA, NASA-GSFC, Greenbelt, Maryland
Ragu Murtugudde
ESCIC, University of Maryland, College Park,
Antonio J. Busalacchi
Laboratory for Hydrospheric Processes, NASA-GSFC, Greenbelt, Maryland
Abstract. A water mass flow formed by the Benguela Current and the South Equatorial Current connects the eastern subtropics to the western tropics through the Atlantic upper thermocline. We perform a process study with an Atlantic OGCM to examine whether and how synthetic subtropical mixed layer heat anomalies would employ this corridor to reach low-latitudes. Our results suggest that it is a realistic scenario and that the time scale and trajectory of the movement can be explained to first approximation by time-mean flow advection, since salinity compensation is tempering the density perturbation. In addition, wave processes seem to influence strongly the evolution of the intensity and the shape of the anomalies. It is apparent that an approach using both perspectives is necessary to fully understand and predict subsurface oceanic teleconnections.
1. Introduction
Recent studies suggest that subsurface equatorward displacement of heat anomalies originating in the subtropics could be one of the mechanisms involved in the decadal variability of the sea surface temperature (SST), particularly in the tropics. Most work along these lines has been in the Pacific Ocean because of possible decadal-scale links to the El Niño Southern Oscillation (e.g., Gu and Philander, 1997). In the Atlantic, subtropical-tropical communication has received less attention. Nevertheless, the decadal variability in this basin, in the tropics (e.g., Servain, 1991) and subtropics (e.g., Tourre et al., 1998), is essential to the climate of the surrounding continents (e.g., Fontaine et al., 1999; Hurrel, 1996) and it is important to look at this potential mechanism. In the South Atlantic below the mixed layer, the South Equatorial Current (SEC) forms a broad thermocline flow, fed by the Benguela Current, that provides the largest part of the shallow water of the tropical and equatorial regions (Stramma and England, 1999). It is therefore an interesting candidate to study as a subsurface pathway to the western low-latitudes for heat anomalies formed in the southeastern subtropics. More generally, the analyzis of propagation of signal along such a pathway is important since it provides insight to the thermocline ventilation mechanism and can be compared to Tritium studies on equator ventilation (e.g., Fine et al., 1987).
Regarding the possibility for heat anomalies to move along subsurface mean flow, Deser et al. (1996) studied one propagation of a subducted mid-latitude temperature anomaly (?0.5ºC) on constant density surfaces (isopycnals) at the speed of the mean currents in the North Pacific. Schneider et al. (1999a) confirmed this finding, regardless of the sign of the anomalies. These authors suggest that perturbations follow the subtropical gyre circulation from around 30ºN to 15ºN-18ºN, whereas equatorward of this latitudes, the physics of heat anomalies is dominated by local wind stress curl. Hence, north of 18ºN at least and as invoked by Gu and Philander (1997), these results can be interpreted as advective processes within the ventilated thermocline. The mechanism responsible for the anomalies to behave like passive tracer could be a density compensation by salt anomalies (Deser et al., 1996; Schneider et al., 1999a,b). A competing theory is that such movements correspond to baroclinic planetary waves within a mean current, having trajectories and speeds comparable (but smaller) to the mean current (Liu, 1999; Huang and Pedlosky, 1999). Liu and Shin (1999) compared the subsurface evolution of a synthetic temperature anomaly and a companion passive tracer in an idealized North Atlantic OGCM. They found comparable pathways, but a propagation of the passive tracer two times faster, and interpreted this discrepancy as resulting from the propagation of a so-called « advective baroclinic mode ».
Several questions are addressed here: how does the subsurface pathways of heat anomalies compare to mean flow advection? How significant are the associated anomalous density signal and currents and what is the importance of salt compensation? What is the role of wave versus advective processes? And at last, are mid-latitudes anomalies capable of traveling several years and reaching the low latitudes in the South Atlantic Ocean as a coherent signal?
2. The numerical model and its pathways
We employ a reduced-gravity, primitive equation OGCM (Gent
and Cane, 1989; Murtugudde et al., 1996) coupled to an advective atmospheric
mixed layer model (AML, Seager et al., 1995). The model domain covers (50ºN,50ºS)x(100ºW,20ºE)
with 20 sigma layers in the vertical and a horizontal resolution of 1/2º.
Note that this is not an isopycnal ocean model. The ocean is forced with
climatological winds (Hellerman and Rosenstein, 1983) and the AML is forced
with climatological precipitation (Oberhuber, 1988), ERBE radiation and
ISCCP cloudiness. The coupled model is spun up to steady state with annual
mean forcing fields in order to study the physics of temperature anomaly
propagation within a stationary