Portal de Periódicos da CAPES

Subsurface connections in the eastern tropical Pacific during La Niña 1999–2001 and El Niño 2002–2003

2011; American Geophysical Union; Volume: 116; Issue: C12 Linguagem: Francês

10.1029/2011jc007624

2156-2202

Ivonne Montès, Wolfgang Schneider, François Colas, Bruno Blanke, Vincent Échevin,

Marine and coastal ecosystems

Journal of Geophysical Research: OceansVolume 116, Issue C12 Free Access Subsurface connections in the eastern tropical Pacific during La Niña 1999–2001 and El Niño 2002–2003 Ivonne Montes, Ivonne Montes ivonnem@udec.cl Programa de Postgrado en Oceanografía, Departamento de Oceanografía, Universidad de Concepción, Concepción, Chile Centro de Investigación Oceanográfica en el Pacífico Sur-Oriental (Fondap-Copas), Universidad de Concepción, Concepción, ChileSearch for more papers by this authorWolfgang Schneider, Wolfgang Schneider Centro de Investigación Oceanográfica en el Pacífico Sur-Oriental (Fondap-Copas), Universidad de Concepción, Concepción, Chile Departamento de Oceanografía, Universidad de Concepción, Concepción, ChileSearch for more papers by this authorFrancois Colas, Francois Colas Institute of Geophysics and Planetary Physics, University of California, Los Angeles, California, USA Laboratoire d'Océanographie et du Climat Expérimentation et Approches Numériques, IRD/IPSL/UPMC, Paris, FranceSearch for more papers by this authorBruno Blanke, Bruno Blanke Laboratoire de Physique des Océans, UMR 6523, CNRS/Ifremer/IRD/UBO, Brest, FranceSearch for more papers by this authorVincent Echevin, Vincent Echevin Laboratoire d'Océanographie et du Climat Expérimentation et Approches Numériques, IRD/IPSL/UPMC, Paris, FranceSearch for more papers by this author Ivonne Montes, Ivonne Montes ivonnem@udec.cl Programa de Postgrado en Oceanografía, Departamento de Oceanografía, Universidad de Concepción, Concepción, Chile Centro de Investigación Oceanográfica en el Pacífico Sur-Oriental (Fondap-Copas), Universidad de Concepción, Concepción, ChileSearch for more papers by this authorWolfgang Schneider, Wolfgang Schneider Centro de Investigación Oceanográfica en el Pacífico Sur-Oriental (Fondap-Copas), Universidad de Concepción, Concepción, Chile Departamento de Oceanografía, Universidad de Concepción, Concepción, ChileSearch for more papers by this authorFrancois Colas, Francois Colas Institute of Geophysics and Planetary Physics, University of California, Los Angeles, California, USA Laboratoire d'Océanographie et du Climat Expérimentation et Approches Numériques, IRD/IPSL/UPMC, Paris, FranceSearch for more papers by this authorBruno Blanke, Bruno Blanke Laboratoire de Physique des Océans, UMR 6523, CNRS/Ifremer/IRD/UBO, Brest, FranceSearch for more papers by this authorVincent Echevin, Vincent Echevin Laboratoire d'Océanographie et du Climat Expérimentation et Approches Numériques, IRD/IPSL/UPMC, Paris, FranceSearch for more papers by this author First published: 16 December 2011 https://doi.org/10.1029/2011JC007624Citations: 33AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract [1] The subsurface connections between the Equatorial Current System (ECS) and the Peru Current System (PCS) between 1999 and 2005 are investigated with a primitive-equation, eddy-resolving regional model that is forced with realistic atmospheric and lateral oceanic conditions. Specific attention is given to the 1999–2000 La Niña and the 2002–2003 El Niño. The model's skill is assessed through a comparison with satellite-derived sea level anomalies and in situ sea surface temperature time series. The model reproduces fairly well the known dynamics of the region for climatological conditions, and the numerical solution obtained for the particular 1999–2000 and 2002–2003 events presents patterns rather typical of cold and warm phases of El Niño-Southern Oscillation (ENSO). Eulerian and Lagrangian diagnoses are used to derive relevant information about the density and velocity vertical structures of the ECS and the PCS. The transports of the major currents in the region are shown to differ a lot between the 1999–2000 La Niña and the 2002–2003 El Niño. The equatorial subsurface currents transfer significantly more water into the eastern tropical Pacific during La Niña than during El Niño, whereas the Peru-Chile Undercurrent (PCUC) carries more water during El Niño. The equatorial subsurface currents, and especially the primary Southern Subsurface Countercurrent, contribute to 80% of the PCUC transport during the 1999–2000 cold phase. This ratio falls down to only 20% during the 2002–2003 warm phase. Key Points We present an interannual solution for the ocean circulation in the ETP Equatorial subsurface currents differ from mean to La Niña to El Niño phases Most of the incoming transport of the ECS recirculates westward during both ENSO events 1. Introduction [2] The El Niño-Southern Oscillation (ENSO) has been widely studied, especially because of its impact on marine biological productivity and the climate of the Americas [Wang and Fiedler, 2006]. This phenomenon is characterized by an irregular interannual oscillation of tropical Pacific upper ocean temperatures and is the strongest climate variation on time scales ranging from a few months to several years [Wang and Picaut, 2004; Latif and Keenlyside, 2009]. [3] Under normal conditions, the trade winds accumulate warm surface water in the western Pacific and draw colder upwelled water to the surface along the equator in the eastern Pacific. Hence, the zonal sea surface temperature (SST) gradient is eastward and the thermocline is pushed down in the west and elevated in the east [McPhaden, 2004; McPhaden et al., 2006]. Variations in the strength of the trade winds generate equatorially trapped baroclinic disturbances at several time scales in the western Pacific. These disturbances propagate eastward in the shape of equatorial Kelvin waves and are held responsible for the transmission of equatorial variability all the way to the South American coast; they propagate eventually poleward as coastal-trapped waves [Enfield and Allen, 1980; Brainard and McLain, 1987]. At the interannual scale, changes in these winds can break down or reinforce the east–west surface temperature contrast and shoal or deepen the equatorial thermocline, altering the propagation of equatorial waves. Furthermore, the variability of the trade winds modifies the circulation patterns at the ocean surface and subsurface [Fiedler et al., 1992; Strub et al., 1998; Blanco et al., 2002; Wang and Fiedler, 2006; Colas et al., 2008]. A remarkable effect is the weakening or even disappearance of the Equatorial Undercurrent (EUC) during the ENSO warm phases (El Niño events), in contrast to its strengthening during the cold phases (La Niña events) [McPhaden and Hayes, 1990; Kessler and McPhaden, 1995; Izumo et al., 2002]. [4] The EUC is one major component of the Equatorial Current System (ECS) that lies beside and in connection with the Peru Current System (PCS). The feeding of the Peru-Chile Undercurrent (PCUC) by the eastward subsurface equatorial currents (EUC and the primary and Southern Subsurface Countercurrent (pSSCC) and secondary Southern Subsurface Countercurrent (sSSCC)) under normal atmospheric and oceanic conditions shows evidence of the ECS-PCS interaction [Lukas, 1986; Johnson and Moore, 1997; McCreary et al., 2002; Montes et al., 2010]. The PCUC is part of the PCS, flows poleward over the shelf slope, and is an important source of productive coastal upwelled waters off Peru owing to its high nutrient content [e.g., Huyer et al., 1987]. [5] Considering how direct and important the connection between these two current systems is, the PCS ought to be strongly affected by the ENSO variability, with occurrence of interannual variability in currents such as the PCUC. Furthermore, sustained advection of warm waters from the ECS toward the Peruvian coast during El Niño events could have dramatic consequences on the regional climate. For example, the accumulation of anomalously warm water off Peru can generate an onshore geostrophic flow that can further delay the coastal upwelling recovery during El Niño [Colas et al., 2008]. [6] Therefore, our main interest here is to investigate how this connection is modified during the cold and warm phases of ENSO. We use a high-resolution numerical model to simulate the ocean circulation in the eastern tropical Pacific (ETP) for the period 1999–2005, and we address the following issues: (1) the impact of ENSO on the subsurface Equatorial Current System and the PCUC in the ETP, (2) the connections between the subsurface Equatorial Current System and the PCUC during contrasted ENSO conditions and (3) the transports associated with these different connections. According to the Niño3.4 index (http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.shtml), our simulation period spans a strong La Niña (August 1999–July 2000), a moderate El Niño (May 2002–February 2003), and a weak El Niño (June 2004–February 2005). Our analysis focuses on the strong cold phase and the moderate warm phase. [7] This paper is structured as follows. The description of the model and its validation are given in section 2. Section 3 provides the Eulerian description of the subsurface ECS and of the PCS and uses a Lagrangian analysis to understand their connections under contrasted ENSO conditions. Our conclusions are presented in section 4. 2. Methodology 2.1. Regional Ocean Model [8] The numerical model employed in this work is the Regional Ocean Modeling System (ROMS). ROMS is a split-explicit, free-surface, primitive-equations ocean model, based on the Boussinesq approximation and hydrostatic vertical momentum balance. A complete description of ROMS is given by Shchepetkin and McWilliams [2005, 2009]. [9] The model domain covers the ETP from 22°S to 4°N and from 94°W to the South American coast (Figure 1). The spatial resolution is 1/9° (1∼12 km) with 32 terrain-following vertical levels (with a higher resolution in the upper ocean layer). The bottom topography is derived from the ETOPO2 (2′ resolution) data set [Smith and Sandwell, 1997]. The model is run from August 1999 to December 2005, driven by wind stress derived from daily QuikSCAT satellite scatterometer data gridded at 1/2° resolution [Liu et al., 1998]. The model is also forced by fresh water and heat fluxes extracted from the Comprehensive Ocean–atmosphere Data Set (COADS) ocean surface monthly climatology at 1/2° resolution [Da Silva et al., 1994]. This data set is also used to restore model SST and sea surface salinity to climatological values through a heat-flux correction [Barnier et al., 1995]. Five-day varying variables from the Simple Ocean Data Assimilation Parallel Ocean Program (SODA-POP) reanalysis version 2.4.3 [Carton and Giese, 2008] at 1/2° resolution are used to force our model at its three open boundaries (north, west, and south). Thus, the interannual signals related to ENSO variability are introduced in the model through its lateral boundaries as a remote oceanic forcing and through the regional wind stress as a local forcing. The SODA representation of August 1999 is also used as initial conditions. The model solution is considered stable after a 3 year spin-up that used the three-times-repeated first year (August 1999–July 2000) of the model forcing. Model outputs are averaged and stored on a daily basis. Preprocessing and postprocessing of the simulation were achieved partially with the use of ROMSTOOLS (http://roms.mpl.ird.fr/) [Penven et al., 2008]. Figure 1Open in figure viewerPowerPoint Model domain, mean surface currents (with one vector every fifth grid point), and SST (color shaded) of the interannual simulation. Current speed is scaled according to the reference vector located in the upper right hand corner. The color bar indicates temperature in °C. The area wherein Lagrangian experiments are carried out is bounded by gray solid lines and by the coast of South America, and the names of the corresponding five control sections are given in blue. 2.2. Lagrangian Algorithm [10] The Lagrangian diagnoses are based on the off-line mass-preserving algorithm ARIANE (http://www.univ-brest.fr/lpo/ariane/) [Blanke and Raynaud, 1997; Blanke et al., 1999], which is used here to calculate trajectories of numerical floats (particles) within the daily archive of our interannual, three-dimensional velocity field. The approach allows the full description of individual trajectories as well as volume transport estimates on the basis of the infinitesimal transport weight allotted to each numerical float and transported without alteration along its trajectory. The volume of water transported from an initial to a final geographical section is computed by simply summing the infinitesimal transport of the numerical floats that achieve the connection that is being considered. [11] For this particular study, we run our Lagrangian diagnoses within the area that spreads from 10°S to 2.5°N and from 92°W to the west coast of South America. The edges of this domain are subdivided into five adjacent sections (Figure 1). The western section is set at 92°W, just west of the Galapagos Island. It strategically intercepts the equatorial currents relevant for our study and is used to define the release positions of all the numerical floats. The four other sections bound the domain north and south and correspond to possible contrasted destinations for the particles. The northern edge is located at 2.5°N and bears a northwestern section and a northeastern section that are delimited by longitudes 92°W and 83°W and by longitude 83°W and the coast of Colombia, respectively. The southern edge is located at 10°S and also bears two sections, a southwestern section and a southeastern section, that are delimited by longitudes 92°W and 82°W and by longitude 82°W and the Peruvian coastline, respectively. [12] We focus on the connection achieved from the western section to the southeastern section since it corresponds to the pathway of particles that travel from the ECS to the nearshore area occupied by the PCUC at 10°S (see section 3 for more details). To do this, millions of particles are initialized on the western section at depths above 400 m and in areas where the eastward zonal velocity exceeds 0.01 m s−1, with the distribution in time and space proposed by Blanke et al. [1999] and with an individual weight, related to the local magnitude of the inflow, not to exceed in our case the maximum value of 10−3 Sv d−1 (with 1 Sverdrup = 106 m3 s−1). All particles are integrated forward in time until they reach one of the five previously defined control sections (see section 3.4 for more details), with a maximum integration time of 1000 days (about 2.75 years) allowed for completing the connection (all the particles exited the control domain within this time interval). The starting dates on the initial section range sequentially from the first day of the model archive (1 August 1999) until 31 July 2002. Insofar as the ocean model was run until the end of December 2005, this sampling strategy does allow all the released particles to complete their integration within the archive available for the interannual simulation. 2.3. Validation [13] To evaluate the realism of our ROMS simulation, we compare the model outputs with three observational data sources: the sea level anomaly (SLA) derived from the monthly satellite data produced by Segment Sol Multimissions d'Altimétrie, d'Orbitographie et de Localization Précise/Data Unification and Altimeter Combination System (SSALTO/DUACS) [Ducet et al., 2000], SST from the monthly Advanced Very High Resolution Radiometer (AVHRR) Pathfinder product (http://www.nodc.noaa.gov/sog/pathfinder4km/) gridded at 4 km resolution, and the monthly SST data from six coastal stations of the Instituto del Mar del Perú (IMARPE, http://www.imarpe.pe). [14] Figure 2 shows the comparison of monthly SLA maps (with respect to the mean sea level over August 1999–December 2005) for our model solution and satellite altimetry for two typical La Niña and El Niño situations, in October 1999 and November 2002, respectively. In accordance with ENSO dynamics [McPhaden, 2004], negative anomalies of ∼6 cm appear in the equatorial region and along the coast at the peak of the La Niña event, whereas an opposite pattern, with maximum positive anomalies of ∼10 cm, is prominent at the peak of the El Niño event, for both the model solution and the observations. The model shows fair agreement with observations for large-scale patterns and amplitudes. The coastal waveguide is in particular well reproduced during warm ENSO conditions. Some differences are of course noticeable, e.g., larger model positive SLAs in the southern half of the domain and a farther southward extension of the equatorial Kelvin wave observed in the model during El Niño. Part of the discrepancy can be attributed to the difference in spatial resolution (the gridded satellite data are available at 1/3° resolution) and to the existence of a blind zone in altimetry coverage near the coast. Figure 2Open in figure viewerPowerPoint Monthly mean SLA maps from the model (Figures 2a and 2c) and AVISO/DUACS satellite (Figures 2b and 2d) altimetry during (a, b) the peak of La Niña (October 1999) and (c, d) El Niño (November 2002). The SLA is color shaded. The color bar indicates the SLA in cm. [15] Figure 3 shows the model and AVHRR-Pathfinder SST for the two same characteristic months and emphasizes the major oceanographic patterns found during La Niña and El Niño events. During La Niña (Figures 3a and 3b), three main features characterize the regional dynamics: the cold upwelled water along the shore of Peru and northern Chile, the meridional SST gradient confined to the equator (mostly known as the equatorial front), and the cold tongue extending northwestward from the coast (∼5°S) to the equator. During El Niño (Figures 3c and 3d), the coastal upwelled water is about 2°C to 3°C warmer than during La Niña and is confined closer to the coast, the equatorial front is less pronounced and is shifted southward, and the cold tongue is replaced by warmer water entering the eastern Pacific from the west. In general, the model reproduces fairly well these features, although it tends to overestimate the upwelling intensity near the coast, especially during La Niña. The wind product in use could explain this since the QuikSCAT data are known to overestimate the strength of the wind field within 50–100 km of the shoreline [Capet et al., 2004; Croquette et al., 2007; Colas et al., 2011]. Over the whole domain, the model bias is of the order of −1°C and −0.9°C during the cold and warm events, respectively. Figure 3Open in figure viewerPowerPoint Monthly mean SST maps from the model (Figures 3a and 3c) and AVHRR-Pathfinder satellite (Figures 3b and 3d) data during (a, b) the peak of La Niña (October 1999) and (c, d) El Niño (November 2002). The SST is color shaded. The color bar indicates the SST in °C. [16] Finally, we compare time series of model monthly SST anomalies (computed with respect to the average value over August 1999–December 2005) with equivalent data derived from IMARPE coastal stations located at Paita (5.06°S), Chiclayo (6.76°S), Chicama (7.72°S), Chimbote (9.05°S), Huacho (11.12°S), and Callao (12.06°S) (Figure 4). Some discrepancies are obvious, such as peaks underestimated by the model (e.g., in March 2002 for Paita, Chimbote, and Huacho). They can be attributed to the model spatial resolution, which is too coarse to resolve finely the nearshore processes, and to overestimation of the atmospheric forcing near the coast, which leads to colder than observed SSTs. Nevertheless, the comparison shows fair agreement between model and observed time series with a high linear correlation coefficient (0.74 to 0.87) except at the southernmost station (0.58). Model and in situ data show both a pronounced seasonal signal, in phase, with peak values in summer. Figure 4Open in figure viewerPowerPoint SST anomaly (°C) time series obtained from the model (solid line) and IMARPE data (dashed line) at six locations along the coast of Peru: Paita (5.06°S), Chiclayo (6.76°S), Chicama (7.72°S), Chimbote (9.05°S), Huacho (11.12°S), and Callao (12.06°S). Observations were available for the entire period of the model simulation (August 1999–December 2005). [17] Furthermore, the mean state of our interannual numerical solution can be compared with the mean state of former ROMS climatological configurations obtained for about the same study area. Our solution closely matches that of Penven et al. [2005] as well as that of Montes et al. [2010] (e.g., SST bias of the order of 0.5°C and 0.3°C, respectively). Both above mentioned studies were forced with QuikSCAT-derived wind fields, but open boundary conditions were obtained from the Ocean Circulation and Climate Advanced Modeling (OCCAM) project [Webb et al., 1997] and SODA model, respectively. 3. Results and Discussion 3.1. An Eulerian View of the Equatorial Current System: Mean Conditions, La Niña, and El Niño [18] In this section, we focus on the vertical structure of the ECS, especially the subsurface zonal currents, during the cold and warm phases of ENSO. Daily model outputs for the most representative months of each event are averaged at 92°W, just west of the Galapagos Islands, over the meridional section that will be used below to release numerical particles. We choose the periods October–December 1999 and October–December 2002 for La Niña and El Niño periods, respectively. Additionally, we use the mean current data of the interannual run for the years 1999–2005 to calculate a reference state. In order to study the behavior of the currents to the east of the Galapagos Islands, we also analyze their vertical structure at 87°W. 3.1.1. Mean Conditions [19] The ECS is formed by four main currents (Figure 5): the westward South Equatorial Current (SEC) at the surface and the eastward EUC and Tsuchiya jets (pSSCC and sSSCC) at the subsurface [e.g., Kessler, 2006]. Vertical sections of modeled zonal velocities at 92°W clearly reproduce these dynamical structures under mean conditions. Near the surface, these currents include the divided lobes of the SEC after passing the Galapagos Islands and flowing westward north and south of the equator in the upper 50 m, with maximum mean velocities of 40–50 cm s−1 (Figure 5a). Eastward flowing currents include the subsurface EUC that extends from 2°S to 2°N and from ∼30 to ∼300 m depth with a maximum mean velocity greater than 30 cm s−1 at 0°, ∼100 m; the pSSCC with a maximum mean velocity core of ∼14 cm s−1 at ∼4°S, 100 m; and the sSSCC with a maximum mean velocity core of ∼3 cm s−1 at ∼7°S, 170 m. At 87°W, before approaching the Galapagos Islands (Figure 5d), the SEC extends over the full meridional extent of the model domain and takes up the upper 50 m, exhibiting a maximum velocity core of ∼30 cm s−1 at 0.5°S. The three equatorial subsurface currents show a slight eastward reduction in their velocity cores as well as in their latitudinal and vertical extents. The EUC core keeps flowing along the equator at ∼100 m depth, at ∼22 cm s−1, i.e., a decline of ∼40% compared with that of the upstream section. The cores of the pSSCC and sSSCC have maximum velocities of ∼11 and ∼2.5 cm s−1, respectively, showing a slight reduction inside the study region. Nevertheless, both currents keep their mean latitudinal and vertical positions at ∼4°S, 100 m, and ∼7°S, 170 m, respectively. Although the available observational data at 92°W are too scarce to allow a thorough comparison with our model outputs, this description with respect to depth, latitude, and associated velocities west of the Galapagos Islands is consistent with other model results [e.g., Donohue et al., 2002] and observational data [e.g., Johnson et al., 2002], as summarized by Kessler [2006]. This confirms the reasonable degree of realism of the model solution. Moreover, the vertical current structure of the ECS averaged between 86°W and 87°W, based on a climatological ROMS simulation and depicted by Montes et al. [2010, Figure 3] agrees with our interannual simulation despite different open boundary conditions and model domains. Insofar as our configuration includes the Galapagos Islands, our results confirm the reformation of the EUC after flowing around the archipelago, as proposed by Steger et al. [1998] and Karnauskas et al. [2007]. Figure 5Open in figure viewerPowerPoint Vertical (0–500 m depth) sections of zonal velocity and density at 92°W (Figures 5a–5c) and 87°W (Figures 5d–5f) during (a, d) mean conditions (September 1999 – December 2005), (b, e) La Niña (September–December 1999), and (c, f) El Niño (September–December 2002) events. The density is color shaded, and the color bar indicates σt units. Solid black (gray) contours indicate eastward (positive) and westward (negative) flows, respectively. White contours mark zero velocity. 3.1.2. ENSO Cold Phase [20] During the warm and cold phases of ENSO, the four currents are also present, but their features in depth and intensity along 92°W and 87°W are notably different from the mean conditions. During the 1999–2000 La Niña, at 92°W (Figure 5b), the SEC still reveals its two lobes, but the northern branch flows faster than the southern one (∼60 cm s−1), with maximum mean velocities of ∼70 cm s−1. The EUC is somewhat narrower (2°S–1°N) and exhibits a much deeper vertical extension (from ∼30 to 500 m depth), and its core is located deeper, near 150 m at the equator with a maximum velocity greater than 35 cm s−1, i.e., a value close to mean conditions. The pSSCC also occupies a wider depth range (between ∼50 and >500 m depth), and its core (3°S–4.6°S) is found slightly deeper at about 120 m near 4°S with maximum mean velocities ∼22 cm s−1, thus stronger than under mean conditions. The core of the sSSCC (6.5°S–8°S) is centered somewhat shallower, close to 150 m at 7.5°S; however, it has a stronger maximum mean velocity of ∼13 cm s−1. In the same way as the other subsurface currents, the sSSCC exhibits its maximum vertical extension during La Niña conditions, with its upper boundary situated very close to the surface and extending down to around 500 m depth. The vertical structure of the currents at 87°W is much different (Figure 5e). Maximum velocities within the SEC still appear north of the equator, but also at the equator (∼20–45 cm s−1) where the SEC occupies the upper 100 m of the water column (instead of ∼50 m at 92°W, after flowing around of the Galapagos Islands). It spreads over the full meridional model domain, as under mean conditions. The EUC is slightly wider (2°S–1.5°N) than at 92°W and is confined to the subsurface between 100 and 400 m depth, with maximum core velocities of ∼32 cm s−1 at ∼220 m, which are significantly deeper than those close to the western boundary of the model domain. The pSSCC as such vanishes at 87°W; only a small remnant subsists farther south (∼5°S) and much closer to the surface than west of the Galapagos Islands, with maximum velocities of ∼7 cm s−1. The sSSCC is also situated farther south (∼8.5°S), with its upper limit at the surface and a maximum velocity core of ∼11.5 cm s−1. Aside from the pSSCC at 87°W, all the subsurface currents are strengthened during the 1999 La Niña compared with mean conditions. 3.1.3. ENSO Warm Phase [21] The vertical structure of the ECS during the 2002–2003 El Niño is also very distinct from mean conditions and contrasts with the structure found for the cold phase of ENSO. At 92°W (Figure 5c), the SEC maintains its typical two-lobe shape with its cores occupying the upper 70 m of the water column north (∼0.5°N) and south (∼1.5°S) of the equator, with velocities of 60 and 30 cm s−1, respectively. The outcrop of the sSSCC between 5° and 7°S breaks its meridional continuity. At the subsurface, the most noticeable feature is the unification of the equatorial subsurface currents that appear mostly like one single current with three different cores. The EUC (1°S–2°N) has a drastically reduced vertical extension (from ∼70 to 200 m depth), its core shifts slightly northward and is located near 120 m depth, thus shallower than during La Niña, with a maximum mean velocity of 45 cm s−1, which is stronger than during mean and La Niña conditions. The pSSCC shifts significantly northward (0.5°S–4°S) and is centered much shallower (close to 50 m depth at 1°S) with mean maximum velocities of 25 cm s−1,similar to La Niña conditions. It shows, however, a much less vertical extension (between 50 and ∼120 m) and a wider latitudinal extension than under mean and La Niña conditions. The core of the sSSCC is found at shallow depths of about 50 m, closer to the equator than during normal or cold conditions, around 6.5°S, with a maximum mean speed of 20 cm s−1, i.e., the fastest configuration within all three scenarios. This current extends from 250 m depth to the surface and thus displays its minimal vertical extension of all three scenarios. Moreover, a westward current known as the Equatorial Intermediate Current (EIC) is visible during this period below the EUC. At 87°W (Figure 5f), the vertical structure of the ECS is basically the same as at 92°W, although the mean velocities and extensions of the currents are changed. The core of the SEC flows with maximum velocities smaller than 25 cm s−1 and is located slightly farther south than at 92°W. Again, its meridional extension is interrupted by the outcrop of the sSSCC, however, to a lesser extent. The EUC and the pSSCC cannot be distinguished from each other anymore. They f