A quasi-synoptic view of the frontal circulation in the Crozet Basin during the Antares-4 cruise

Abstract

With the major objective of estimating the secondary vertical circulation, a fine-grid survey of hydrographic and current fields across the Crozet Basin frontal zone was made as a physical component of the multidisciplinary Antares-4 cruise. This frontal zone forms the boundary of the Antarctic Circumpolar Current and the Agulhas Return Current, with the quasi-permanent existence of strong eddy activity. Prior to any quantitative diagnosis of the vertical circulation from the fine-grid data, it is necessary to place the mesoscale circulation and its time variability of the survey area within the basin-wide context. We give a quasi-synoptic view of the ambient frontal circulation during the cruise period, using satellite-derived chlorophyll concentrations and sea-surface temperature. Particular emphasis is placed on evaluation of the degree of synopticity of our fine-grid measurements by comparing mesoscale features on satellite images. The limited survey area (1.5° latitude×2° longitude) lies precisely within the southernmost and easternmost convergent zone of the two current systems, cutting across mesoscale meanders having typical wavelengths of 400 km and periods of 4 months. Both these satellite images and TOPEX/POSEIDON sea-surface height data indicate little temporal variability of the survey area during the 14 days of measurement. Although fortuitous, this quasi-steady state of circulation enables us to use the ensemble of fine-grid data as a set of synoptic data for diagnosing the vertical circulation using quasi-geostrophic dynamics. Preliminary analysis of the horizontal and vertical hydrographic structure of the fine-grid hints at the existence of wave-like cross-frontal intrusions having a mean along-front wavelength of about 30 km.

Résumé

Avec l’objectif majeur d’estimer la circulation verticale secondaire, des mesures de champs hydrologiques et courantologiques sur une grille fine placée sur la zone frontale du bassin de Crozet ont été récemment effectuées dans le cadre de la composante physique de la campagne multidisciplinaire d’Antares-4. Cette zone frontale constitue la frontière entre le Courant Circumpolaire Antarctique et le Courant de Retour des Aiguilles, avec l’existence quasi-permanente de fortes activités tourbillonnaires. Avant un quelconque diagnostic quantitatif de la circulation verticale à l’aide des données de la grille fine, il est nécessaire de replacer adéquatement la circulation méso-échelle et la variabilité temporelle de la zone dans le contexte du bassin entier. Nous présentons une vue quasi-synoptique de la circulation frontale pendant la campagne, à l’aide des données satellitales de concentration chlorophyllienne et de température de la surface de la mer. Un effort particulier a porté sur l’évaluation du degré de synopicité de nos mesures dans la grille fine, en comparant à un mois d’intervalle les traits méso-échelles déterminés à partir des images satellitales. La zone d’étude, d’une dimension limitée (1.5° latitude×2° longitude), s’est révélée se trouver précisément dans la zone de convergence la plus au sud et la plus à l’est des deux systèmes de courants, en coupant les méandres de longueur d’onde caractéristique 400 km et de période 4 mois. L’ensemble de ces images satellitales et des données TOPEX/POSEIDON sur les hauteurs de la mer indique une faible variabilité temporelle de la zone d’étude durant ces 14 jours de mesures. Bien que fortuit, cet état quasi-constant de la circulation nous permet d’utiliser l’ensemble des données de la grille fine comme un jeu de données synoptiques pour le diagostic de la circulation verticale à l’aide de la dynamique quasi-géostrophique. Une analyse préliminaire des structures horizontales et verticales de la grille fine met en évidence des intrusions quasi-ondulatoires à travers la zone frontale, avec une longueur d’onde moyenne le long des fronts d’environ 30 km.

Introduction

The subtropical frontal zone in the southwest Indian Ocean recently has attracted particular attention from the biogeochemical community because of its enhanced primary productivity and associated oceanic CO₂ sink. Weeks and Shillington (1994) showed from coastal-zone color scanner data that the frontal zone south of Africa between the Agulhas Return Current (ARC) and Subtropical Front (STF) exhibits relatively high levels of chlorophyll (1.5 mg m⁻³), which are several times larger than typical Southern Ocean values. Their Plate 4 hints at the eastward extension of a zonal band of high concentrations of chlorophyll into the Crozet Basin. Because of the geostrophic equilibrium of the field of mass, the juxtaposition of warm saline subtropical waters carried by the ARC with colder fresh subantarctic waters of the Antarctic Circumpolar Current (ACC) makes the boundary of these two water regimes in the Crozet Basin one of the most intensified frontal zones of the Southern Ocean. The large-scale circulation of the basin and associated strong eddy activity caused by topographically generated powerful meanders and eddies have been well documented (Park, 1990; Park et al (1991), Park et al (1993); Park and Saint-Guily, 1992; Park and Gambéroni, 1995). An elevated oceanic CO₂ sink has been frequently observed north of the ACC (Poisson et al (1993), Poisson et al. (1994)), which may be attributed to the juxtaposition of the cooling effect on warm subtropical waters and biological utilization effect on nutrient-rich subantarctic waters (Takahashi et al., 2000). It also has been suggested that the high productivity along oceanic fronts is related to enhanced nutrients supplied from below by the secondary vertical circulation (Allanson et al., 1981). In fact, oceanic fronts are known as areas of intensified vertical circulation, with alternating upwelling and downwelling cells occurring across the fronts (Pollard and Regier, 1992; Allen et al., 1994; Rudnick, 1996).

It is thus clear that precise knowledge of the 3-D circulation is indispensable for a better understanding of the coupling between the physical and biogeochemical components of a frontal system. An approach to this knowledge was attempted during the Antares-4 cruise, which was carried out across the frontal zone of the central Crozet Basin in January–February 1999 on R/V Marion Dufresne II. This multidisciplinary cruise, part of a French contribution to the Southern Ocean-Joint Global Flux Study (SO-JGOFS), had the general objective of documenting, understanding, and modelling the biogenic fluxes and processes controlling the CO₂ pump across the frontal zone (Denis et al., unpublished document, 1998). The physical component of the cruise, which is the main concern here, aimed to resolve the density and current fields of the frontal zone as finely as possible in order to estimate the vertical circulation using quasi-geostrophic dynamics or omega equations. To this end, second-order partial derivatives of density and geostrophic currents are necessary (see Pollard and Regier, 1992, for details). Therefore, accomplishment of the physical task demands the following two critical but necessarily conflicting constraints: the synopticity of measurements and their fine resolution in space. Failure to satisfy these constraints adequately will render hopeless any effort to determine the vertical component of currents (Allen et al., 2001). The pre-cruise hypothesis was that the dominant mesoscale features of the area have a typical wavelength of about 400 km and a period of 4 months. A trade-off between fine spatial resolution for the omega equations and synopticity led us to carry out the ensemble of measurements within 14 days along a fine-grid bounded by a small parallelogram of 1.5° latitude×2° longitude, parallel to the TOPEX/POSEIDON (T/P) ascending track 129 (Fig. 1). This fine-grid was composed of 9 SW–NE lines, with a zonal interval of 20 km. The easternmost line, corresponding to the T/P track 129, was surveyed by seven full-depth conductivity–temperature–depth (CTD) casts, 33 km apart. This was followed for the rest of the grid by continuous hydrographic measurements on passage to a depth of 1200 m using the TOWYO (Prieur and Sournia, 1994), CTD sensors towed by the ship and connected to an automatic winch, permitting continuous up and down casts while the ship travelled at a reduced speed. With a nominal ship's speed of 4 knots, the along-track TOWYO resolution was about 6 km on average. Throughout the cruise, the 75 kHz acoustic Doppler current profiler (ADCP) installed on the Marion Dufresne II was operated. Combined with the ship's global positioning system (GPS) navigation and ship's heading referenced to 3-D GPS, currents down to 500 m were recorded with errors of <2 cm⁻¹ (Pollard et al., unpublished document, 1999). This estimate does not include noise introduced by tides or inertial motions, which is difficult to estimate from our one-time survey. Such motions, where they exist, should be eliminated as much as possible when we estimate ADCP geostrophic currents by an objective analysis, which has not been attempted in this descriptive paper.

Before pursuing any dynamic calculations from the fine-grid data, it is necessary: (1) to position the cruise area adequately within the context of basin-wide mesoscale eddy activity, and (2) to verify whether the observed frontal structures within the fine-grid satisfy the imposed constraints (synopticity and fine resolution), crucial conditions in applying quasi-geostrophic dynamics. The first step is also intended to validate the circulation schematic (see Fig. 1) that had been prepared during the cruise using all available in situ data gathered in a limited geographical region. This will be achieved based on comparison with satellite-derived chlorophyll concentrations and sea-surface temperature (SST). Temporal variations of these satellite images together with those of altimeter-derived sea-surface height will be used to gauge the degree of synopticity of our fine-grid data. Detailed maps of property distributions on a chosen depth or isopycnal, together with those on selected cross-front or along-front sections of the fine-grid, may hint at the dominant horizontal and vertical scales that could be related to the frontal ageostrophic circulation, thus to the secondary vertical circulation. Finally, we are concerned also with the spatial resolution of our fine-grid, in particular, the orientation of grid lines relative to the current field. The choice of grid lines parallel to the T/P track 129 relied on our pre-cruise knowledge of the large-scale circulation pattern of the region, suggesting a frontal zone running roughly perpendicular to the chosen satellite track. If such is the case, the hydrographic section along the altimetric track will serve as a reference for calibrating altimetric data, thus for monitoring the time evolution of geostrophic transport across the section.

As can be seen later, however, the real situation did not match this presumption and some grid lines were found to lie nearly parallel to the fronts. This is because of the encounter with a highly meandering current system whose major orientation was undetermined until the survey advanced too far to rectify its initial planning. We will discuss later on the implication of these “uncommonly” aligned lines for discovering submesoscale frontal waves along the fronts. In summary, this paper is intended to provide a first description of the physical setting that will be useful for conducting further quantitative studies of frontal dynamics using the Antares-4 data, and will provide the physical context for biological and chemical studies.