JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 101, NO. C1, PAGES 923-939, JANUARY 15, 1996 Intensive surveysof the Azores Front 1. Tracers and dynamics Daniel L. Rudnick Scripps Institution of Oceanography, La Jolla,California JamesR. Luytcn WoodsHole Oceanographic Institution, WoodsHole, Massachusetts Abstract. The hypothesis that fronts are sites of active subduction is examined using density, temperature, salinity, and horizontal velocity data from a trio of surveys of the AzoresFront done in May 1991and March 1992.Thesesurveys were made using a SeaSoar equipped with a conductivity-temperature-depth profilerand a shipboard acoustic Doppler current profiler. The potential density andpotential vorticity indicate that dense water from the north sideof the front may be sliding downbeneath the surface outcrop. This apparently subducting isopycnal hasa greatdeal of temperature and salinity variability. Horizontal velocity is nearly parallel to isopycnals, indicating that the time rate of change and verticaladvection mustbe small. The thermalwind balance is observed to be valid,especially in the region of the largest horizontal density gradients. Shear at the base of the mixedlayeris likelydue to near-inertial motions. The potential vorticity is dominated by the planetary vorticity, except at the front,where vertical shears (the tilting term) become large. The tiltingterm acts to reduce the magnitude of the potential vorticity at the front, in agreement with simple theoretical models. The magnitude of the tilting term is similar to the total vorticity in the seasonal thermocline. 1. Introduction The similarityin propertiesof water masses in the upper thermocline to late winter surface waters has been well estab- lished. Iselin [1939] first noted the similarity in the tempera- ture-salinity (T-S) curves for north-south surface sections and verticalprofiles in the subtropical gyre.He further suggested the possibility that March surface water moves alongisopyc- nals to set the T-S relationship in the thermocline. Stommel [1979] proposed a mechanism (theso-called Ekman demon) by which water from the base of the year'sdeepest mixedlayer in March is then swept into the interiorby Ekman pumping in a region of Sverdrup dynamics. A relatively simple modelof this process, neglecting mixed-layer dynamics, was presented by Luyten eta!. [1983].The rate of thisprocess, sometimes called subduction, is set by the Ekman pumping vertical velocityat order 10 -6 m s -• [Isemer and Hasse, 1987]. What is the mechanism by which the water escapes the mixed layer? Usingthe concept of an entrainment velocity [de Szoeke, 1980], Nurser and Marshall[1991] assumed continuity across the mixed-layer base and determinedthe "subduction velocity" (the negative of the entrainment velocity) to be pro- portional to the net heat flux. The apparentresult is that subduction is possible onlyin the presence of net surface heat- ing. It is reasonable to askif subduction might be possible in the absence of any heating. The hypothesis of thisstudy isthat oceanic fronts are sites of active subduction. Two essential characteristics of fronts lead to this hypothesis. First, fronts are often sites of surface con- Copyright 1996by the AmericanGeophysical Union. Paper number 95JC02867. 0148-0227/96/95JC-02867505.00 vergence whichmay, to some extent, be balanced by localized downwelling. Second, because of largehorizontal density gra- dients, fronts are regions of enhanced vertical shearand pos- siblystrong verticalmixing. The thrusthere is to present evi- dence of the first characteristic. While it will be shown that shear is increasedat fronts, no direct observations of mixing are presented. We investigate subduction at frontsthrough an intensive description of a series of fronts and a careful exam- ination of frontal dynamics. The importance of vertical circulationat fronts has been demonstrated by a number of studies. Pollard andRegier [1990, 1992] have examined vertical circulation throughtracersand by the solution of the omegaequation. These studies of the subtropical front in the western North Atlantic have inferred vertical velocities asstrong as40 m d -•. Rudnick and Weller [1993]have calculated the vertical velocity necessary to close the heat budgetusinga moored array of current meters and have found similarmagnitudes. The goal of this paper is to demonstrate the likelihood of strong vertical velocities by ex- amining tracer fields. The quantification of vertical velocity will be addressed in part 2 [Rudnick, 1996]. We investigate the hypothesis that fronts are sitesof sub- duction with a trio of surveys of the AzoresFront, one in May 1991 and two in March 1992. The surveys were conducted using a SeaSoarequippedwith a conductivity-temperature- depth(CTD) profilerand a shipboard acoustic Dopplercur- rent profiler(ADCP), A summary of the surveys and data is givenin section 2. Results are presented in verticalsections across the front and horizontalmaps; the mapping procedure is described in section 3. The hydrographic variables are pre- sented in section 4. The velocity, dynamics, and potentialvor- ticity are discussed in section 5. Finally, section 6 provides a summary and conclusions. 923