The mesoscale variability in the Caribbean Sea. Part II: Energy sources
Introduction
Previous observations based and model based studies did not permit to draw a clear view of which processes set the characteristics of the eddy field in the Caribbean Sea, but rather showed several points of controversy. The reason is perhaps that many processes can potentially drive the variability and explain the enhanced eddy production in the region. There is neither a general agreement on whether the energetic eddies originate from local processes such as windstress, current instability or topographic effects or from remote processes such as propagation of Atlantic Rossby waves or advection of North Brazil Current (NBC) rings.
Among the local processes, the presence of a strong mean windstress curl (WSC, see Fig. 2) in the Central Caribbean has been proposed to be a source of energy for the Caribbean mesoscale activity. A model study which excludes incoming perturbations from the Atlantic shows eddies produced southwest of Hispaniola (Oey et al., 2003). The authors suggest that the local mean WSC is responsible for this generation, although they admit that the instability of the Caribbean Current could be inadequately simulated in their experiments. Andrade and Barton (2000) associate the enhancement of eddy activity in the central Caribbean during July-October with the maximum curl of the North Trade Wind during this period. However they do not consider the influence of NBC rings and that the mean Caribbean Current may also have annual variability correlated with eddy production. Recent observations suggest local wind generation is not enough to explain the eddies vertical structure. The real eddies observed by Silander (2005) in the Venezuela Basin are deep and energetic (maximum swirl speed ranged from 0.3 to 0.6 m s−1): it is unlikely that pure wind stress curl, which is not particularly strong in the Venezuela Basin, can spun up in just few weeks at these latitudes such energetic eddies reaching more than 1000 m depth. In addition, altimetry data (e.g., Guerrero et al., 2004) show that most of the anticyclones occur in the southern part of the Caribbean Sea where the WSC is mostly cyclonic.
Another local process which can be responsible for eddy generation and growth is the instability of the main Caribbean Current. Conversion from Mean Kinetic Energy (MKE) to Mean Eddy Kinetic Energy (MEKE) has been proposed to contribute to eddy growth (Carton and Chao, 1999) but not demonstrated. Other hypothesis, provided by Andrade and Barton (2000) is that unstable meandering of the Caribbean Current could form some eddies. Richardson (2005) also suggests that instabilities of the anticyclonic shear of Caribbean jets could help the incoming NBC rings vorticity to organize and amplify into energetic eddies. Diagnostics of our simulations discussed below in Section 2 demonstrate that the main Caribbean currents are prone to be unstable and calculations of energy conversion terms between mean flow (mean kinetic energy and mean stratification) and eddies confirm that the two fields exchange energy. Although the eddy field can transfer energy to the mean flow in some particular regions (Yucatan and Nicaraguan coasts), conversions of energy mainly benefit the eddy field. In addition, we noticed a close correspondence between regions of maximum conversion and regions where MEKE increases. Some regions are dominated by barotropic instability (e.g., the Cayman Sea), whereas other regions are dominated by baroclinic instability (Colombia and Venezuela Basins).
Topography and geography of the Caribbean Sea are quite complex (see Fig. 1 in Jouanno et al., 2008) and they might also have some local influence on both growth and decay of the eddy field. The main Caribbean Current (which we refer to here as the southern Caribbean Current, sCC) flows along the continental coast; our simulations show that the shape of the coast line and the narrowness of the Lesser Antilles passages accelerate the current and modify its instability properties. Topography and geography can also dissipate energy and destroy coherent structures. Altimetry data (Andrade and Barton, 2000) and simulations (Carton and Chao, 1999) show that most of the eastern Caribbean eddies seem dissipated by topographic features in the coastal waters of Nicaragua and do not reach the Yucatan Channel. Our simulations suggest that the strong decrease of MEKE along the Nicaraguan coast in association with an acceleration of the mean flow through the narrow Chibcha Channel allows the Cayman Basin to develop variability with its own particular characteristics.
Concerning a possible remote origin of the variability, the process most often proposed in the literature is the advection of NBC rings through the Lesser Antilles. They originate in the equator, where the reflection of long Rossby waves on the Brazilian coast generates a cyclonic-anticyclonic system which travels northwestward along this coast. Nonlinear interactions with the coast and the -effect contribute to the growth of anticyclonic eddies (Jochum and Rizzoli, 2003) which can reach the narrow and shallow passages of the Lesser Antilles (Fratantoni et al., 1995). Carton and Chao (1999) suggest that the interaction of the incoming NBC rings with the topography around the Island of Trinidad and Tobago, results in pairs of cyclones anticyclones which become part and interact with the Caribbean Current. Murphy et al. (1999) also invoke the NBC retroflection and suggest that potential vorticity (PV) of the NBC rings is advected through the Lesser Antilles, and acts as finite amplitude perturbations for mixed barotropic and internal mode baroclinic instabilities which result in mesoscale features. Analyzing drifter data, Richardson (2005) proposes that anticyclones in the Venezuela Basin could originate with the vorticity advected by the NBC rings, hypothesis consistent with process studies on eddy flux through islands chains (Simmons and Nof, 2002, Tanabe and Cenedese, 2008). Nevertheless, drifters show few direct evidences of rings entering completely. Another process study demonstrates that Atlantic Rossby Waves can pass through ocean barriers such as the Lesser Antilles (Pedlosky, 2000).
Recently, Chérubin and Richardson (2007) proposed a link between the number of eddies in the eastern Caribbean, inferred from drifters, and the presence of the fresh water plume, which would enhance the potential vorticity gradients during August-December. It is interesting to contrast their results with surface geostrophic velocity anomalies derived from 15 years of altimetry data (AVISO). They indicate that the maximum of eddy energy in the eastern Caribbean occurs during June-July (see Fig. 1). Clearly, it is difficult to relate the arrival of the freshwater plume in the eastern Caribbean during August–December with the local maximum of eddy energy two months before and also use it as the source of energetic eddies all along the year. The arrival of the freshwater plume can certainly impact eddy generation and maintenance (e.g during the period September-November) but given the characteristics of the instability processes in the model and altimetry observations, it is difficult to make it the main source of variability in the eastern Caribbean.
So there is no consensus on what are the energy sources of the Caribbean eddies and questions remain open:
- •
What is the dominant process in the genesis of the most energetic Caribbean eddies?
- •
Which processes are responsible for the westward intensification of the eddy energy in the Caribbean Sea?
- •
What is the real impact of the NBC rings on the Caribbean eddy variability?
The aim of this paper is to try to answer these questions. The different numerical experiments used to solve this problem, as well as the characteristics of the Caribbean mean flow and eddy field, are described in part I of this study (Jouanno et al., 2008). As shown in part I, the most energetic mesoscale variability is embedded or occurs in the sCC, so we mainly focus on this current. We start by analyzing its conditions of instability in Section 2 and by examining in Section 3 the exchange between mean and eddy field along its core. In Section 4 we discuss how the various Caribbean inflows organize into an unstable current. Section 5 details and illustrates the formation process of the most energetic eddies in the eastern Caribbean. Section 6 investigates the importance that NBC rings and Atlantic perturbations could have in setting the characteristics of the Caribbean variability. Section 7 analyzes the dynamics of other regions of eddy production, such as the northern Caribbean Current (nCC), the Panama-Colombia Gyre or the Cayman Current. Finally, Section 8 gives a summary and conclusions.
Section snippets
Instability conditions of the sCC
The sCC has a strong vertical shear and its flow is narrow and strong, as shown by model (e.g., part I of this study) or observations (Hernández-Guerra and Joyce, 2000). The geographical position, intermediate between the equator and high latitudes, allows the Caribbean Basin to support relatively fast baroclinic and barotropic responses. All these characteristics are favorable to the growth of baroclinic instability particularly for a westward current (Talley, 1983). Strong horizontal shear
Eddy-mean flow interactions in the sCC
Energy analysis gives determining information on the generation of the mesoscale activity by the mean circulation and highlights the strong dynamical differences between different regions of the Venezuela and Colombia Basins. We focus on the interactions between energy components. Effects of advection, forcing or dissipation are not considered here. Following the framework described in Beckmann et al. (1994), the energy transfer terms are given by:
The origin of the PV contrast
The sCC constitutes a reservoir of available energy, and part of this energy is released through instability processes to the perturbations. This reservoir has to be recharged or maintained in some way. As explained by Fig. 16 of Jouanno et al. (2008), the sCC belongs to various large scale systems and is fed by: (1) the upper branch of the MOC, as a flux of light water masses (these waters have crossed the equator and were taken to the Lesser Antilles by the Guyana Current and the NBC rings),
Formation of the most energetic Caribbean eddies
Analysis of velocity snapshots and instantaneous QGPV field is illuminating to better understand how the largest Caribbean eddies originate in the eastern Caribbean Sea. Fig. 9 shows snapshots of the QGPV every 20 days at 30 m depth (left column) and 150 m depth (right column). Instantaneous velocity fields at the same depths are superimposed. At day 0, an anticyclone (pattern referenced as A on the figure) and its cyclonic counterpart are growing, respectively, on the northern and southern side
Influence of the NBC rings
Results in the previous sections indicate with confidence that the mean Caribbean Currents are unstable and that their instability increases the variability in the Caribbean Basin. It does not mean that instability of the currents drives all the Caribbean mesoscale variability. As mentioned in the introduction, several works link the Caribbean eddy production with NBC rings (e.g., Carton and Chao, 1999, Murphy et al., 1999, Richardson, 2005). In this section, we compare three experiments
Other regions of eddy production and growth
The eddies embedded in the sCC are the most energetic of the Caribbean Sea, as discussed previously and shown by the high MEKE values in the southern part of the Venezuela and Colombia Basins (Fig. 12 in Jouanno et al., 2008). Out of the sCC pathways, other mesoscale perturbations exist. They are sufficiently strong to impact or dominate the local variability, and are not directly linked with the sCC eddies. In this section, we analyze the dynamics of three regions for which both model and
Conclusions
The present work has established a coherent framework for understanding the previously reported heterogeneity of eddy activity and conflicting ideas of eddy generation. It was shown that the instability of the main Caribbean currents is the most relevant explanation for the growth of the mesoscale eddies in the Caribbean Sea. Indeed, any meridional section in the Caribbean Sea satisfies the necessary condition for instability in the quasi-geostrophic approximation, i.e. the meridional gradient
Acknowledgements
We acknowledge the provision of supercomputing facilities by the Institut pour le Développement des Ressources en Informantique Scientifique of the Centre National de la Recherche Scientifique, by the Departamento de Supercomputo of the UNAM and by the CICESE (Conacyt project 621-259). Some computations were also performed at Bakliz of DGSCA, UNAM. Altimetry data were produced by Salto/Duacs and distributed by Aviso (http://www.jason.oceanobs.com), with support from CNES. The fine grid/coarse
References (31)
- et al.
Caribbean current variability and the influence of the amazon and orinoco freshwater plumes
Deep See Research I
(2007) - et al.
The mesoscale variability in the caribbean sea. Part I: simulations with an embedded model and characteristics
Ocean Modell
(2008) Caribbean Current and eddies as observed by surface drifters
Deep Sea Research II
(2005)- et al.
Eddy development and motion in the Caribbean Sea
Journal of Geophysical Research
(2000) - et al.
On the seasonal variability and eddies in the North Brazil Current: Insights from model inter-comparison experiments
Progress in Oceanography
(2001) - et al.
On the generation and role of eddy variability in the central North Atlantic Ocean
Journal of Geophysical Research
(1994) Baroclinic instability and the short wave cut-off in terms of potential vorticity
Q.J.R. Meteorol. Soc.
(1966)- et al.
Caribbean Sea eddies inferred from TOPEX/Poseidon altimetry and a 1/6 Atlantic ocean model circulation
Journal of Geophysical Research
(1999) - et al.
Geographical variability of the first-baroclinic Rossby radius of deformation
Journal of Physical Oceanography
(1998) - et al.
The interaction of a baroclinic mean flow with long Rossby waves
Journal of Physical Oceanography
(2005)
The Evolution and demise of North Brazil Current rings
Journal of Physical Oceanography
Rings of the North Brazil Current: their strucure and behaviour inferred from observations and a numerical simulation
Journal of Geophysical Research
Atmosphere Ocean Dynamics
Tracking eddies in the Caribbean Sea using the A VISO altimetry analysis
Eos, Transactions AGU, Western Pacific Geophysics Meeting Supplement
Cited by (34)
Investigation of ocean environmental variables and their variations associated with major Loop Current eddy-shedding events in the Gulf of Mexico
2024, Deep-Sea Research Part II: Topical Studies in OceanographyEnergetics of eddy–mean flow interactions in the deep western boundary current off the northeastern coast of Brazil
2023, Deep-Sea Research Part I: Oceanographic Research PapersTriggering mechanisms of the Yucatan upwelling
2023, Continental Shelf ResearchCitation Excerpt :The impact of eddies and vorticity perturbations go beyond the Yucatan Channel and influence the Loop Current extension and its anticyclonic eddy releases (Candela et al., 2003; Athié et al., 2012; Huang et al., 2013; Mildner et al., 2013; Androulidakis et al., 2021). Eddies can also be generated locally from instability of the main current (Jouanno et al., 2009; Sheinbaum et al., 2016; Yang et al., 2020). However, most studies have focused on the Loop Current area and their impact remains to be determined.
ENSO Modulates Mean Currents and Mesoscale Eddies in the Caribbean Sea
2023, Geophysical Research LettersAtmosphere-Ocean Processes Governing Inflow to the Northern Caribbean Sea
2023, Journal of Marine Science and Engineering