Subthermocline variability in the South Eastern Pacific: interannual to decadal timescales

Get Complete Project Material File(s) Now! »

Wind-driven circulation

In each ocean basin, the large-scale circulation over the first hundreds of meters organizes around a great anticyclonic gyre, and in the case of the Pacific and Atlantic oceans, these features are roughly symmetric about the equator. Researchers in the mid-1800s started to realize that the currents changes in the upper ocean followed a change in the wind field by a matter of hours, and the hypothesis suggesting that the frictional stress of the wind was the responsible of such relationship was first proposed by Croll4 in 1875. From that period on, the intrinsic turbulent nature of the natural water bodies became known and the concept of turbulent or “eddy” viscosity was developed. Making use of this concept, following works developed the major elements in wind-driven circulation theory (e.g. how the rotation of the earth is responsible for the deflection of wind-driven currents (Ekman, 1902; Nansen, 1898), the equatorial surface currents system and its relationship with the trade winds (Sverdrup, 1947), the solutions of the wind-driven circulation using a realistic wind field (Munk, 1950)).
One of the main particularities of the large-scale ocean circulation is the so-called westward intensification, as shown in figure 1.1, where the flow lines tend to be close together over the western boundary, while they are spaced in the eastern boundary, illustrating that the flow is swift in the west, while it is sluggish on the opposite side of the basin. This feature characterizes all ocean basins and was first explained by Stommel (1948).
In that work, the author based his demonstration in simplified, theoretical models of the ocean and wind pattern, which consisted in a constant-depth rectangular ocean on one hemisphere, and a wind field only varying in the latitudinal direction, as show in figure 1.2. Stommel’s work exposed for the first time that the variation of the Corio-lis parameter with latitude is responsible for the western intensification of the currents in the ocean, and following works explained this feature in terms of vorticity conser-vation (e.g. Pond and Pickard, 1983; Rhines, 1986; Lozier and Riser, 1989). In general terms, the vorticity put into the ocean by the wind stress must be taken out (or bal-anced) by friction. In the west, a strong friction is needed to balance out the vorticity acquired with the poleward increase in f, and for this strong friction to occur strong currents with strong shear are needed. As a result of the western intensification, we find slow currents with low lateral shear in the eastern side of the basin. This induces a poor ventilation of the circulation (Luyten et al., 1983) thus the “age” of the water masses found in the eastern ocean boundaries is higher compared with what is found in the western side (Fig. 1.3), as it is the case when comparing the eastern and western boundaries of the south Pacific.

The Humboldt Currents system

The south basin of the Pacific Ocean (60◦S, 150◦E-70◦W) accounts for 25% of the total oceans volume, and hosts in its eastern boundary one of the four major coastal upwelling systems of the world. Named after the German naturalist Alexander von Humboldt, the Humboldt Currents System (HCS) extends from southern Chile (near 45◦S) to northern Peru (~4◦S), and from the coast to ~90◦W. This currents system is most notable for its prodigious production of small pelagic fish, which represents 27% of the current annual landing for the fisheries in the Pacific Ocean and has played a key role in the development of several countries for decades.
Figure 1.2: Flow patterns (streamlines) for a simplified wind-driven circulation model in the northern hemisphere with: (a) Constant Coriolis force, (b) Coriolis force increasing linearly with latitude. Idealized wind stress pattern used to force the model is also shown. After Stommel (1948).

Large scale circulation in the HCS

The large-scale oceanic circulation that shapes the HCS is closely related to the trade winds, which are dynamically set up in both the southern and northern hemispheres by the pressure gradient developed between a low pressure area near the equator and a high pressure region localized around 30◦. This pressure gradient induces an air-flow from the mid-latitudes towards the equator, and creates a zonally-narrow wind convergence zone, known as the Intertropical Convergence Zone (ITCZ; Fig. 1.4). In the Pacific, the ITCZ is located on average around 10◦N, although its position varies seasonally in connection with the easterlies. Another convection zone characteristic of the tropical Pacific is the South Pacific Convergence Zone (SPCZ), which corresponds to a low-level convection band associated with a subtropical maximum in cloudiness, precipitation and sea surface temperature (Kiladis et al., 1989). It extends from the southeast Pacific (30◦S-120◦W) to Papua New Guinea, where it merges with the ITCZ (Fig. 1.4). Both the ITCZ and the SPCZ determine the large-scale mean rainfall pattern tropical Pacific band (Takahashi and Battisti, 2007a,b).
The variations of the wind system in the HCS are influenced by the latitudinal shifts of the ITCZ and trade winds in the northern hemisphere, and the latitudinal variations of the SPCZ (Karoly et al., 1998), but they are mainly driven by the shifts of the South Pacific High (SPH) present off central Chile (~30◦S). The changes in both the position and the intensity of the SPH impact the wind field in the HCS (Rutllant et al., 2004) and couple to the orographic effect of the Andean mountain range, which results in nearly alongshore equatorward winds close to the coast (Fig. 1.4).
Figure 1.4: Mean sea level pressure (blue to orange contours) and wind magnitude and direc-tion (arrows) in the south Pacific for the period 2000-2008 (ICOADS dataset). Black contours correspond to mean rainfall values of 3 and 6 mm day−1, evidencing the low-level convergence bands. Dashed red line qualitatively marks the mean position of the ITCZ.
HCS. Off Peru, the equatorward winds are intense and almost year-round, with a max-imum during austral winter (Bakun and Nelson, 1991; Dewitte et al., 2011). On the other hand, the wind seasonal cycle off northern Chile peaks during austral spring (Blanco et al., 2002), and during austral summer off central/southern Chile (Garreaud and Muñoz, 2005), which has been related to the seasonal migration of the SPH (An-capichún and Garcés-Vargas, 2015). Off central Chile, the atmospheric conditions are also subject to the excitation of low atmospheric pressure systems that are trapped to the coast by the pressure gradient between the marine boundary layer and the coastal orography, and propagate polewards (Garreaud et al., 2002).
In the ocean, the circulation in the HCS is composed by several equatorward and poleward alternating currents, with a main surface equatorward flow north of 45◦S (Strub et al., 1998) located next to the eastern rim of the SPH (Fig. 1.5). At around 10◦S, the main flow turns offshore and flows into the South Equatorial Current (SEC), and a weak ramification continues equatorward and joins the SEC at around 8◦S (Wyrtki, 1966). Below the surface, the ramifications of the eastward flowing Equatorial Under Current (EUC) nourish the subsurface poleward components of the HCS: (1) the Peru-Chile Under Current (PCUC), a subsurface poleward flow found over the slope and outer shelf off the Peruvian and Chilean coasts and (2) the Peru-Chile Counter Cur-rent (PCCC), located between 150 and 300 Km offshore (Strub et al., 1998; Fig. 1.5). In relationship with its equatorial origins, the PCUC has a distinctive hydrographic sig-nature, characterized as relatively saltier, higher in nutrients and lower in oxygen than the surrounding waters, and these characteristics have allowed to trace it as far south as 48◦S (Silva and Neshyba, 1979).
The HCS exhibits an important seasonality in relation to the yearly cycle of the envi-ronmental forcing. On a regional scale, the oceanic response to environmental forcing at seasonal timescale is primarily related to the annual net insolation cycle (Takahashi, 2005), although ocean dynamics contribute to make the system heterogeneous and in-duce distinctive responses to forcing off Peru and Chile, which has been interpreted as the result of a compensation between large-scale dynamical signals (Dewitte et al., 2008a). On the other hand, the oceanic response to environmental forcing next to the coast is closely related to the wind forcing and manifests as coastal upwelling, which is the process responsible for the high primary productivity observed along the coast in the HCS. This coastal upwelling is sustained by the alongshore equatorward wind stress that generates an Ekman divergence of the currents next to the coast, which is in turn compensated by a vertical upward flow of nutrient-rich waters carried by the PCUC (Kelly and Blanco, 1984; Wyrtki, 1963). In addition, the large scale along-shore wind stress decreases over a few hundred kilometers next to the coast, due to the coastal orography, the surface drag gradient between land and sea, and the air-sea interactions over cool seawater (see Capet et al., 2004). It is expected that this phenomenon, also known as wind “drop-off”, would create an onshore wind stress gradient which would in turn result in a negative wind stress curl and ultimately an Ekman pumping (Bakun and Nelson, 1991) that could also contribute to the vertical upwelling.
Although the mechanisms behind the coastal upwelling in the HCS are related to the large scale wind system present in the region, the small scale variations in the ocean circulation and coastal orography contribute to make the system heterogeneous, encompassing three well-defined upwelling subsystems along the HCS (Montecino and Lange, 2009): (1) the year-round and highly productive upwelling system off Peru, (2) a low productivity “upwelling shadow zone” in southern Peru and northern Chile, and (3) a productive and seasonal upwelling system in central-southern Chile.

The south Pacific Oxygen Minimum Zone

In addition to a highly productive upwelling system, the SEP encompasses the southern portion of one of the most extensive Oxygen Minimum Zones of the planet (OMZs; Paulmier and Ruiz-Pino, 2009). The OMZs are regions in the ocean character-ized by extremely low concentrations of Dissolved Oxygen (DO) in the water column (DO < 60 µM), as a result of complex interactions between the ocean circulation and the biogeochemical cycles (Fig. 1.6; Karstensen et al., 2008). In the HCS, the intense biological production that takes place in the euphotic zone of the water column (first 200m) is accompanied by an important subsurface remineralization of organic matter, which translates as a significant DO demand in the mesopelagic zone (Capone and Hutchins, 2013). This important DO demand couples to poor ventilation, related to a nearly stagnant circulation (Luyten et al., 1983), which allows the OMZ to persist in time. Among the most relevant impacts of the OMZ in the SEP we find the habi-tat compression of the organisms, given that the OMZ represents a respiratory barrier (Prince and Goodyear, 2006). Additionally, the biogeochemical cycles that take place at extremely low DO concentrations are involved in the local production of climatically-active gases, such as CO2 (Paulmier et al., 2011) and N2O (Kock et al., 2016), which are then outgassed to the atmosphere. In this sense, the OMZ has an impact on both the local ecosystems and on the global climate.
Despite its potential implications, several questions remain regarding the OMZ variability and long-term trends (Stramma et al., 2010). Recent light has been shed upon the mechanisms that shape the OMZ (Bettencourt et al., 2015), emphasizing the central role of the mesoscale structures. However, it has not been clarified yet whether or not these structures influence the variability of the OMZ and its long-term evolu-tion, considering that the mesoscale activity appears as a conspicuous feature in the SEP circulation.

READ  A model for telecommunication technology transfer /diffusion into rural areas

Mesoscale features

Over the large-scale circulation pattern found in the HCS, rich mesoscale variabil-ity in the form of eddies or vortices, filaments and squirts is superimposed (Fig. 1.7). In the HCS, eddies are characterized by a radius between 50-150 km (Chaigneau et al., 2008; Chaigneau and Pizarro, 2005b) and can persist for months, traveling thou-sands of kilometers offshore from their genesis region. Eddies are mainly generated from baroclinic instabilities induced by the vertical shear of the currents near the coast (Leth and Shaffer, 2001), and propagate westward with translation velocities between 3-7 cm s−1 (Chaigneau and Pizarro, 2005b). Observations have shown that the HCS is highly populated by eddies between 9◦S and ~36◦S, and they participate in the heat (Colas et al., 2012) and salt balance between the offshore and coastal waters through lat-eral fluxes, even exceeding the mean advective current fluxes in the coastal upwelling region (Chaigneau and Pizarro, 2005a). Despite this evidence, the net contribution of eddies to the SEP’s heat budget is still in debate, due to conflicting modeling and obser-vational results (cf. Mechoso et al., 2014). Recent observational analyses even suggest that eddies would not substantially contribute to the surface layer heat budget in the offshore SEP (Holte et al., 2013).
In the SEP, the mesoscale structures also induce a coupling between physical and biogeochemical processes (McGillicuddy et al., 1998), which significantly extends the high primary production zone associated to the coastal upwelling while moving off-shore. Observational evidence off central Chile (29◦-39◦S) indicates that eddies are responsible for more than 50% of the winter chlorophyll-a peak in the offshore coastal transition zone (Correa-Ramirez et al., 2007). In this sense, eddies might represent a pathway that links the highly productive coastal upwelling region with the (essentially oligotrophic) offshore waters. Nevertheless, this vision is still on debate regarding the highly productive HCS and recent results have established that in fact eddies might contribute to reduce primary productivity as a result of the transport of nutrients from the nearshore to the open ocean (Gruber et al., 2011).

Teleconnection with the equatorial Pacific

Aside from the local atmospheric and oceanic forcing, the variability in the HCS is intimately related to the variability that takes place in the equatorial Pacific, at timescales ranging from intraseasonal to seasonal (Pizarro et al., 2002; Shaffer et al., 1999), and from interannual (Pizarro et al., 2001, 2002; Vega et al., 2003) to interdecadal (Monte-cinos et al., 2007). The equatorial region behaves as a waveguide, and allows for the (b) 17/01/2014, (c) 13/09/2014 and (d) 06/07/2014. The coastal orography and clouds correspond to corrected reflectance (true color), also acquired by MODIS. Data source: zonal propagation of different types of waves. One of the most prominent modes of variability corresponds to the Intraseasonal Equatorial Kelvin Wave (IEKW), gener-ated in the central equatorial Pacific by intraseasonal westerly wind pulses. The IEKW travels to the east and impinges on the American continent. As shown by idealized models, part of the energy is reflected by the eastern boundary and travels to the west as a long Rossby wave, and part is deflected and travels poleward (Cane and Sarachik, 1977; Moore and Philander, 1977) as a free Coastal Trapped Wave (CTW). The South American coast behaves then as an extension of the equatorial waveguide. In contrast to the IEKW, the phase speed of the CTW strongly depends on the shape of the con-tinental slope (Brink, 1982; Clarke and Ahmed, 1999) and stratification (Allen, 1975), and its vertical structure varies with latitude. As the wave travels polewards, the inter-nal Rossby radius of deformation decreases and the bottom topography becomes more important in determining the vertical structure of the gravest baroclinic modes (Brink, 1980). In this manner, at low latitudes the CTW structure is essentially baroclinic, how-ever, as latitude increases, the dominant vertical structure tends to be barotropic (Brink, 1982).
As they travel polewards, the CTWs induce perturbations in the density and pres-sure field over the continental shelf and slope, which has permitted to observe pole-ward propagating signals along the South American coast at a wide range of frequen-cies. Fluctuations of sea level and currents off Peru in the synoptic band (period ~10 days) have been reported to have little relationship with the local wind (Cornejo-Rodriguez and Enfield, 1987; Enfield et al., 1987; R. L. Smith, 1978). Rather, they re-late to the poleward propagation of first baroclinic mode CTWs (Brink, 1982; Romea and R. L. Smith, 1983), forced by synoptic-scale mixed Rossby-gravity waves (or Yanai waves), propagating eastward in the equatorial Pacific (Clarke, 1983; Enfield et al., 1987). Further studies unveiled the relationship between the sea level and currents fluctuations observed in the HCS at intraseasonal frequencies (30-90 days) and the re-mote influence through first baroclinic mode CTWs (Hormazábal et al., 2002; Illig et al., 2014; Shaffer et al., 1997; Spillane et al., 1987), forced by baroclinic IEKWs impinging on the South American coast. Such wind-forced equatorial waves tend to be most promi-nent during austral summer in the central tropical Pacific (Illig et al., 2014; Kessler et al., 1995), in connection with atmospheric convection events that propagate from the central Indian Ocean into the western Pacific.
The IEKW activity is strongly modulated at interannual timescales reflecting the occurrence of El Niño events (Dewitte et al., 2008a; Kessler et al., 1995) and its diversity (Gushchina and Dewitte, 2012; Mosquera-Vásquez et al., 2014), which corresponds to the most prominent mode of climatic variability in the equatorial Pacific.

A brief description of El Niñ

The occurrence of El Niño events entails important disruptions in the tropical Pa-cific climate system primarily associated with changes in the SST gradients and the ocean-atmosphere feedbacks. Due to the efficient teleconnection that links the trop-ical Pacific and the SEP, this mode of tropical climatic variability induces significant upheavals in the SEP climate system, which highlights the fact that the influence of these events extends well beyond the equatorial Pacific. In the present section, a brief description of El Niño is made, and the modulation of the SEP variability induced by these events is presented in the following section.
AVERAGE CONDITIONS: in the tropical Pacific, the wind regime dominated by the easterlies induces a flow to the western part of the basin. As the water flows westwards along the equatorial region, the strong insolation induces a surface warming, and a strong zonal SST gradient is developed (around 12◦C between the western and eastern Pacific). This zonal gradient reflects as a tongue-shaped cold Sea Surface Temperature (SST) that extends from east to west, and a deeper equatorial thermocline develops in the western Pacific (Fig. 1.8a). The warm water found in the western boundary (known as the “warm pool”) generates an important atmospheric thermal convection. The warm and humid air over the warm pool ascends and is driven by the high tropospheric circulation towards the eastern part of the basin, where it loses heat and humidity and settles above the colder sea surface. The air is then driven towards the west by the easter-lies, closing the convective loop. This atmospheric circulation cell is known as the Walker circulation (Fig. 1.8a), in honor of Sir Gilbert Walker, who first de-scribed the strong inverse correlation between the high and low pressure records obtained at the eastern and western Pacific, and coined the term Southern Oscil-lation to describe the “seesaw” behavior of the east-west pressure gradient.
EL NIÑO EVENTS: the onset of an El Niño event is characterized by a relaxation of the easterlies, and a consequent reduction of the equatorial and coastal up-welling, which allows the warm pool to grow and expand eastward in the equa-torial Pacific (Fig. 1.8b). The shift of the warm pool position induces a zonal displacement of the convective Walker cell which in turns changes the evapora-tion patterns, generating droughts in the western part of the basin, and strong precipitations in the eastern part (Fig. 1.8b). These events might be followed by cold events in the central equatorial Pacific, known as La Niña events (Fig. 1.8c), which are considered as an enhancement of the average conditions. Dur-ing a La Niña event, the easterlies are stronger than the average conditions, the warm pool is warmer than average and the waters in the eastern Pacific are even colder (Fig. 1.8c), which couples to a stronger-than-average Walker circulation. This marked ocean/atmosphere coupling between the anomalies generated by El Niño events and the Southern Oscillation is referred to as El Niño-Southern Oscillation (ENSO).

The ENSO arrival at the Humboldt Currents System

Although the ENSO events peak in the equatorial Pacific, wave dynamics carry the anomalous ENSO signal to the SEP. The equatorial Pacific is heavily disrupted dur-ing ENSO events, and this reflects on the IEKW activity. Kessler et al. (1995) showed that during the onset of El Niño events, the eastward extension of the convection cell (associated with a warmer SST) gives the westerlies more fetch to blow upon, gener-ating a more intense IEKW activity. This is coupled to a deepening of the thermocline along the equatorial Pacific, which favors the leading baroclinic mode and enhances the phase speed of the IEKW (Benestad et al., 2002). These changes translate as a mod-ulation of the IEKW at interannual timescale (Dewitte et al., 2008a), which ultimately impacts the CTW activity in the HCS (Enfield et al., 1987). In this sense, Shaffer et al. (1997) documented the intraseasonal currents variations at central Chile (30◦S) dur-ing the 1991-1992 El Niño event, showing that the most important fluctuations were related to the passage of free CTWs arriving from the north.

Table of contents :

1 Introduction 
1.1 Wind-driven circulation
1.2 The Humboldt Currents system
1.2.1 Large scale circulation in the HCS
1.2.2 The south Pacific Oxygen Minimum Zone
1.2.3 Mesoscale features
1.3 Teleconnection with the equatorial Pacific
1.3.1 A brief description of El Niño
1.3.2 The ENSO arrival at the Humboldt Currents System
1.3.3 The connection between the equatorial Pacific and the deep eastern Pacific
1.4 ENSO diversity
1.4.1 ENSO diversity trend
1.5 Thesis motivations and objectives
1.5.1 Scientific objectives and manuscript plan Introduction (français)
2 Methodology and Observations 
2.1 The regional ocean modeling system: ROMS
2.1.1 Coordinate transformation
2.1.2 Pressure gradient errors
2.1.3 Spatial discretization and time stepping
2.1.4 Advection scheme and mixing parametrization
2.2 Long Rossby waves
2.2.1 At the origin of long Rossby waves: the β-plane
2.2.2 The dispersion relation for long Rossby waves
2.2.3 Theoretical approach for vertically propagating Rossby waves
2.2.4 Rossby wave energy flux
2.3 ENSO indices
2.4 Observations
2.4.1 Climatologies and reanalyses
2.4.2 Remote sensing data
3 Subthermocline variability in the South Eastern Pacific: interannual to decadal timescales
3.1 Overview
3.2 Is it possible to use ARGO data to observe the long Rossby wave?
3.2.1 Argo data set
3.2.2 The equatorial Rossby wave
3.2.3 The Extra-Tropical Rossby wave
3.2.4 Conclusion
3.3 Vertical energy flux at interannual to decadal timescales
3.4 Synthesis
4 Ventilation of the South Pacific oxygen minimum zone: the role of the ETRW and mesoscale 
4.1 Overview
4.2 Seasonal variability of the oxygen minimum zone
4.3 Synthesis
5 Conclusions and Perspectives 
Conclusions et Perspectives (français)


Related Posts