SciELO - Scientific Electronic Library Online

Home Pagelista alfabética de revistas  

Servicios Personalizados




Links relacionados


Gayana (Concepción)

versión impresa ISSN 0717-652Xversión On-line ISSN 0717-6538

Gayana (Concepc.) v.68 n.2 supl.TIProc Concepción  2004 


Gayana 68(2) supl. t.I. Proc. : 102-107, 2004 ISSN 0717-652X



Alexis Chaigneau & Oscar Pizarro

PROFC-Universidad de Concepcion, CHILE. Contact:


Based on 9 years (1993-2001) of Sea Level Anomaly measurements from TP/ERS altimeters, we characterize the mesoscale structures of the eastern South Pacific (10-35°S, 70-100°W). Integral zonal, meridional, and temporal scales of the mesoscale activity are described geographically in the study region. The length scales increase toward the north in agreement with the increase of the Rossby radius, whereas timescales show higher values in the south of the domain. Typical values of (Lx, Ly) are around 50 km at 35°S and 80-90 km at 10°S with elongation in the zonal direction revealing the anisotropic nature of the mesoscale structures. Timescales are in the range 15-45 days.

We also estimate a space-varying eddy transfer coefficient K which parameterize the effect of the eddies on the lateral tracer transports. K increases in the regions of high levels of eddy kinetic energy and is in the range of 1107 to 4107 cm2.s-1. Considering mean climatological property gradients, we show that near the coast, the mesoscale eddies can play an important role on the long term heat and salt budgets.



The circulation and the variability of the eastern South Pacific dynamics has been studied based on altimetry measurements (Strub et al., 1995), on historical observations (Pizarro, 1999; Blanco et al., 2001; Shaffer et al., 1997, 1999; Strub et al., 1998), and using numerical models (Leth and Shaffer, 2001). Recently, Chaigneau and Pizarro (2004) based on drifter measurements, have given a new insight on the surface circulation and the turbulent flow in this region of the world ocean (10-35°S, 70-100°W). In particular, these authors have examined for the first time, the typical Lagrangian scales of the mesoscale structures. However, the regular spatial and temporal coverage of the altimetry data offers a unique opportunity to complement this area of investigation.

Eddies are known to be an important mechanism for the transport of heat, salt and momentum in the ocean. Their parameterization in numerical models is an actual topic of scientific debates. In order to determine the role of the mesoscale eddies on the horizontal temperature and salinity transports, we determine a space-varying diffusion coefficient from satellite altimetry data.

This study, based on sea level altimetry measurements describes the mesoscale characteristics (length and timescales) observed by satellite altimetry, and explores the role of the mesoscale activity on the temperature and salinity fluxes. The obtained results provides a useful tool for a comparison with previous studies but overall for the validation of both regional and global models in the eastern South Pacific.


Two major data set are used in this study: the altimetry measurements and hydrographic data from the World Ocean Atlas 2001 (WOA) climatology. Additionally, surface satellite tracked drifters are used to provide the large scale circulation of the study region extending from 10°S to 35°S and from 70°W to 100°W (Figure 1a). Figure 1a shows that the surface circulation inferred from this data set is mainly composed of: the relatively weak eastward South Pacific Current encountered south of 32°S; the strong Chile-Peru Current flowing equatorward east of 84°W; and the westward South Equatorial Current north of 26°S (Chaigneau and Pizarro, 2004).

Monthly temperature and salinity climatology on a ° latitude ° longitude grid from the World Ocean Atlas (WOA, 2001) were used to estimate surface geostrophic velocities relative to 1750 m depth. These geostrophic velocities, combined with surface Ekman's velocities derived from ERS1-2 satellite wind stress, give a good estimation of the surface currents (Chaigneau and Pizarro, 2004). This mean surface circulation U=(U,V) and the WOA surface temperature and salinity gradients (, ) were used to estimate the horizontal large scale heat and salt advections.

Figure 1: a) Large scale surface circulation obtained with drifter measurements [Chaigneau and Pizarro, 2004]. b) Integral timescale of the altimetry measurements. c) Integral zonal and meridional length scales against the Rossby radii. Bold lines represent the linear relations discussed in the text. c) Surface eddy diffusivity.

The satellite altimetry data set used here, is the gridded product of Topex/Poseidon, ERS-1/2, and Jason-1 sea level anomalies (SLA) provided by Archiving Validation and Interpretation of Satellite Data in Oceanography (AVISO). This data set spans the November 1992 ­ May 2002 period with weekly SLA distributed on a 1/3° Mercator Grid. The methods used to process the data and reduced the errors are described in Le Traon et al. (1995, 1998) and in Ducet et al. (2000). Zonal and meridional components of the residual sea surface velocity components were calculated from the SLA assuming the geostrophic relation:

where g is the acceleration due to gravity, f is the Coriolis parameter, and x and y are the eastward and northward directions.

Time mean eddy kinetic energy (per unit of mass), EKE, is calculated using:

where overbar denotes time average.

As in the others parts of the world ocean, the SLA temporal variation in the study region is dominated by the seasonal cycle. Furthermore, the eastern South Pacific is well known for the presence of the El Niño/La Niña phenomena which also affect the SLA, with period of 3-7 years. As we are interested by the mesoscale energetic structures, we filter these low frequency signals by removing at each time the mean SLA over the domain, and the trends in both the zonal and meridional directions. This process decreases the large scale standard deviation of the SLA undesirable for the purpose of this study. For simplicity convenience, the term SLA used above will refer to the filtered data.

In order to describe the turbulent mesoscale flow, we provide the integral length and timescale of the altimetry SLA measurements. The eddy timescale T is calculated from the integration of the SLA temporal autocorrelation function Rt. Due to the presence of numerous negative lobes on the autocorrelation functions, we prefer the quadratic form [e.g., Stammer, 1997]:

where T1 is one-half of the length of the study period. In contrast, the eddy length scales (Lx, Ly) are obtained by the integration of the spatial autocorrelation function (Rx, Ry) in both the zonal and meridional directions:

where L1x and L1y are chosen to be 5° in the zonal and meridional directions respectively.

The role of the mesoscale eddies on the tracer transport is well parameterized with a diffusion coefficient K, via a classic Fickian law [Green, 1970; Stammer, 1998; Jayne and Marotzke , 2002]:

Stammer (1998), calculates a space-varying diffusivity K over the world ocean on a 5 geographical grid, using alongtrack Topex Poseidon altimeter data over the 1993-1995 period. Here we review this method, to provide K over the particular study domain of the eastern South Pacific.

According to Green [1970], the diffusivity can be related to the eddy kinetic energy using:

where á is a correlation coefficient that determine the efficiency of individual eddies to mix tracer particles, and is evaluated to 0.05 [Stammer, 1998]; Tbc is the timescale associated with mixing of tracers

by large scale eddies given by , where Ri is the Richardson number and f the Coriolis parameter. represents the ratio between thermal and mechanic production of turbulent kinetic energy, and is calculated from the WOA (2001) climatology density field and averaged over the 50-1000m water column.

Assuming that the baroclinic instability is the primary eddy source, Tbc can be compared with the eddy timescale T obtained from altimetry. In average over the study domain, Talt = 1.5Tbc, not far from the rate of 2 obtained by Stammer [1998] over the world ocean. Finally, the surface space-varying horizontal diffusivity is calculated only using the altimetry measurements and the relation:

The next section, containing the results is organized as follows: we first describe the characteristics (T, Lx, Ly, and K) of the mesoscale eddies observed by altimetry data in the study region, and then we examined the respective role of the large scale horizontal advection and the lateral turbulent diffusion on the tracer transport in the surface layer. This second part focuses on the following terms of the temperature and salinity evolution equation:



Figure 1b shows that the integral timescale Talt of the mesoscale eddies varies in the study region between 10 days in the north (10-12°S) to around 40 days in the south (33-35°S).

Figure 1c shows the scatter diagram of the zonal and meridional length scales (Lx, Ly) against the Rossby radii of the first boroclinic mode. These Rossby radii were extracted from the 1°x1° climatological atlas of Chelton et al. [1998]. Both the integral length scales increase northward with typical value of 50-60 km at 35°S to 80-90 km at 10°S. North of 15°S these values are always lower than the given Rossby radii. Averaged over a 5° grid in the world ocean, Stammer [1997] found the following linear relation:

where L and R are the Topex-Poseïdon length scale and the Rossby radius respectively.

In the study region, we also observe a linear relation for length scales shorter than 70-80km, but in our case, the coefficients are:

For length scales higher than 70-80 km, Figure 1c indicates different regimes. North of around 23S, the linear relations become:

This suggests that north of 23°S, Lx is higher than Ly of around 5-10km revealing the anisotropic nature of the mesoscale structures, elongated in the zonal direction. This confirms the results of Chaigneau and Pizarro (2004), who also found elongated turbulent flow in the zonal direction, based on satellite tracked drifter measurements.

The diffusivity indicates higher values near the coast and in the south-west of the domain (Figure 1d). This is consistent with observed enhanced values of EKE in these regions from both the altimetry data and the buoy drifting measurements (Chaigneau and Pizarro, 2004). These regions also correspond to higher eddy timescales (Figure 1b). Typical values of diffusivity are in the range 1-4107 cm2 s-1, in agreement with the diffusivities of around 2.5107 cm2 s-1 and 5107 cm2 s-1 found by Chaigneau and Pizarro (2004) in the whole study region with drifter data.

The space-varying diffusion coefficient (Figure 1d) is used to estimate the horizontal temperature and salinity transports by the mesoscale eddies in the ocean surface layer (50m). In the whole domain the large scale circulation tend to decrease the temperature and the salinity of the surface layer of the domain, advecting cooler and fresher water from the south. In average over the region, the respective rates are of around -10 W m-2 and -50 kg m-2 day-1 [Chaigneau and Pizarro, 2004]. Comparatively, the diffusive transports determined from altimetry play a negligible role on the temperature and salinity changes, with respective values of 1.3 W m-2 and 1.7 kg m-2 day-1. However, an average of these fluxes between the continental coast and 3° offshore (Figure 2) suggests that the turbulent transport could play an important role on the long term heat budget in this regio n. Effectively, the turbulent heat flux increases in average to 7 W m-2 near the coast, compare to an average horizontal advection of around ­15 W m-2. Furthermore, the lateral heat and salt fluxes are of the same order of magnitude than the large scale advective fluxes in the 17-25S latitude band due to a strong divergence of the property gradients in this zone (Chaigneau and Pizarro, 2004).


The first goal of this study was to characterize the mesoscale structures in the eastern South Pacific (10-35°S, 70-100°W).We have shown that the length scales increase toward the north in agreement with the Rossby radius changes, whereas the mesoscale timescales vary from 10 days at 10S to 30-40 days at 35S. It also appeared that south of 23S the mesoscale structures are quasi isotropic, whereas north of 23S they are anistropic with a zonal elongation of order of 5-10 km. This confirm the results of Chaigneau and Pizarro (2004) who have shown, based on drifter measurements, that in the whole study region, the turbulent flow has typical length scales of around 40 km and 30 km in the zonal and meridional directions respectively. The obtained timescales are also in agreement with the value of 20-30 days found by Stammer (1997) based on only 3 years of Topex Poseidon measurements.

The altimeter measurements have been used to provide a space-varying eddy transfer coefficient in the study region. The obtained diffusivity show values of 1000-4000 m2 s-1, in the order of magnitude of the results of Chaigneau and Pizarro (2004). Stammer (1998) show typical value of 250 m2 s-1, a factor 10 smaller linked to its estimation of the diffusion coefficient the first 1000 m of the ocean. Finally, this study have shown the importance of the mesoscale structures in providing laterally heat and salt into the surface coastal water. Thus, the mesoscale turbulent flow can play an important role on the long term heat and salt budget of the surface layer along the eastern South Pacific upwelling region.

Figure 2: Mean horizontal heat (a) and salt (b) fluxes averaged between the coast and 3° offshore, as a function of latitude.


The altimetry data were provided by AVISO/CLS. AC was supported by ECOS-SUD and Fundación Andes grant D-13615. This work was also supported by the Chilean National Research Council (FONDAP-COPAS).


Blanco, J. L., A. C. Thomas, M.-E. Carr, & P.T. Strub (2001), Seasonal climatology of hydrographic conditions in the upwelling region off northern Chile. Journal of Geophysical Research, 106, 11451-11467. [         [ Links ]1]

Chaigneau, A. & O. Pizarro (2004), Mean surface circulation and mesoscale turbulent flow characteristics in the eastern South Pacific, from satellite tracked drifters. Journal of Geophysical Research (submitted). [         [ Links ]2]

Chelton, D. B., R. A. de Szoeke, M. G. Schlax, K. El Naggar & N. Siwertz, 1998: Geographical variability of the first-baroclinic Rossby radius of deformation. Journal of Physical Oceanography, 28, 433-460. [         [ Links ]3]

Ducet, N., P. Y. Le Traon, & G. Reverdin (2000), Global high-resolution mapping of ocean circulation from TOPEX/Poseidon and ERS-1 and ­2, Journal of Geophysical Research, 105 (C8), 19477-19498. [         [ Links ]4]

Green, J. S. A. (1970). Transfer properties of the large scale eddies & the general circulation of the atmosphere. Quart. J. Roy. Met. Soc., 96, 157-185. [         [ Links ]5]

Jayne, S. R., & J. Marotzke, 2002. The Oceanic Eddy Heat Transport. Journal of Physical Oceanography, 32, 3328-3345. [         [ Links ]6]

Leth, O. & G. Schaffer (2001), A numerical study of seasonal variability in the circulation off central Chile, Journal of Geophysical Research, 106, 22,229-22,248. [         [ Links ]7]

Le Traon, P.-Y., P. Gaspar, F. Ogor, & J. Dorandeu (1995), Satellites work in tandem to improve accuracy of data, Eos Trans. AGU, 76, 385-389. [         [ Links ]8]

Le Traon, P.-Y., F. Nadal, F. & N. Ducet (1998), An improved mapping method of multi-satellite altimeter data, Journal of Atmospheric and Oceanic Technologies, 15, 522-534. [         [ Links ]9]

Pizarro, O. (1999), Low frequency fluctuations in the eastern boundary current off South America: remote and local forcing, Ph.D. Thesis, Göteborg University, Sweden. [         [ Links ]10]

Shaffer, G., O. Pizarro, L. Djurfeldt, S. Salinas, & J. Rutlant (1997), Circulation and low frequency variability near the Chile coast: remotely-forced fluctuations during the 1991-1992 El Niño, Journal of Physical Oceanography, 27, 217-235. [         [ Links ]11]

Shaffer, G., S. Hormazabal, O. Pizarro, & S. Salinas (1999),Seasonal and interannual variability of currents and temperature off central Chile, Journal of Geophysical Research, 104, 29951-29961. [         [ Links ]12]

Stammer, D. (1997), Global characteristics of ocean variability estimated from regional TOPEX/POSEIDON altimeter measurements, Journal of Physical Oceanography, 27, 1743-1769. [         [ Links ]13]

Stammer, D. (1998), On eddy characteristics, eddy transports, and mean flow properties. Journal of Physical Oceanography, 28, 727-739. [         [ Links ]14]

Strub, P. T., J. M. Mesías, & C. James (1995), Altimeter observations of the Peru-Chile Countercurrent, Geophysical Research Letters, 22 (3), 211-214. [         [ Links ]15]

Strub, P. T., J. M. Mesias, V. Montecino, J. Rutllant, & S. Salinas (1998), Coastal ocean circulation off western South America, The Sea, Vol 11, A. R. Robinson and K. H. Brink Eds, Wiley, 273-313. [         [ Links ]16]


Creative Commons License Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons