inct AmbTropic

Ambientes Marinhos Tropicais: Heterogeneidade Espaço-Temporal e Respostas à Mudanças Climáticas

Apresentação

As mudanças climáticas deverão afetar as características físicas, biológicas e biogeoquímicas das zonas costeiras e oceanos, modificando sua estrutura ecológica, suas funções e os diferentes serviços prestados ao Homem. Estas mudanças tem o potencial de causar sérios impactos sócio-econômicos nas escalas local (center parcs), regional (plataforma e mares rasos) e global (oceano).

As respostas dos ambientes marinhos às mudanças climáticas irão também depender da variabilidade natural destes sistemas e de outras mudanças introduzidas pelo homem como resultado dos diferentes usos dos recursos marinhos.

O bem-estar das comunidades humanas depende intrinsecamente da disponibilidade dos serviços que os ecossistemas costeiros e marinhos provêem. Isto é particularmente importante para a região norte e nordeste, que apresenta em alguns dos seus municípios costeiros, algumas das densidades populacionais mais elevadas do Brasil.

O INCT AmbTropic tem como objetivo central unificador a avaliação de como a heterogeneidade espaço-temporal dos ambientes marinhos tropicais poderá determinar os padrões de resposta destes ambientes e sua resiliência às mudanças climáticas que afetarão o norte-nordeste do Brasil neste século. Este aspecto é de grande importância estratégica para a região.

Outros objetivos deste INCT incluem:

(i) Gerar uma sólida base conceitual sobre os processos, a dinâmica e o funcionamento da zona costeira, plataforma e oceano tropicais do Brasil e sua variabilidade espaço-temporal.

(ii) Dotar a região norte/nordeste do Brasil de uma infraestrutura de pesquisa moderna e adequada para o avanço e consolidação das atividades em Ciências do Mar nesta região.

(iii) Expandir e consolidar a incipiente estrutura de ensino, treinamento e formação de pessoal, para fazer frente ao conjunto de desafios atuais e futuros a serem enfrentados por esta região principalmente como conseqüência das mudanças climáticas.

(iv) Criar uma rede de excelência para dar suporte na solução de problemas prementes que afetam os ambientes marinhos da região norte e nordeste.

(v) Criar mecanismos que possibilitem a transferência ativa de conhecimentos para os principais atores sociais de maneira descentralizada além de garantir acesso irrestrito a todos os dados e informações pretéritas e a serem gerados, pelo INCT.

O INCT AmbTropic é uma iniciativa conjuntas das Universidades Federais da Bahia e de Pernambuco e engloba aproximadamente 200 cientistas distribuídos em mais de 20 instituições de ensino e pesquisa que compartilham esta visão. Os recursos para sua implantação são oriundos do CNPq, FAPESB e CAPES.

O INCT AmbTropic é coordenado pelos professores José Maria Landim Dominguez (UFBA) e Sigrid Neumann Leitão (UFPE)

A sede do INCT AmbTropic está hospedada no Instituto de Geociências da UFBA.

Grandes Linhas

Como já foi apontado pelo IPCC as mudanças globais no clima deverão afetar as características físicas, biológicas e biogeoquímicas das zonas costeiras e oceanos, modificando sua estrutura ecológica, suas funções e os diferentes serviços prestados ao Homem. Estas mudanças tem o potencial de causar sérios impactos sócio-econômicos nas escalas local (zona costeira), regional (plataforma e mares rasos) e global (oceano). Neste contexto, mudanças nas escala local e regional podem ser até mesmo mais relevantes para os Estados costeiros quando comparadas àquelas em escala global.

A região norte e nordeste do Brasil, devido às suas características, apresenta uma oportunidade única para se avaliar de que maneira a heterogeneidade espacial e temporal dos ambientes marinhos tropicais influencia os padrões de resposta destes ambientes e sua resiliência às mudanças climáticas que afetarão a região neste século. Nesta região, compreendida entre os estados do Espírito Santo e Amapá, encontram-se as principais construções recifais do oceano Atlântico Sul Ocidental, os principais deltas brasileiros, uma das áreas mais extensas de manguezais do mundo, uma plataforma continental que varia da mais estreita a mais larga do Brasil, as principais ilhas e montes submarinos, variações extremas nos fluxos de sedimentos e nutrientes, além da sua importância inegável na transferência de calor e massa interhemisférica. Heterogeneidade é assim a característica singular e ao mesmo tempo integradora do INCT AmbTropic. Três escalas espaciais de abordagem estão contempladas:

Zona Costeira (Local) – é uma área de grande heterogeneidade física e biológica e a interface de interação entre as forçantes naturais e antropogênicas.

Plataforma Continental (Regional) – é uma área também de grande heterogeneidade, pouco compreendida e cada vez mais intensamente utilizada pelo Homem. É também uma área que oferece um contexto para interpretação de mudanças e tendências observadas na escala local.

Oceano (Global) – é um componente integral do sistema Terra influenciado por transporte de massa e por suas interações com a atmosfera.

Cada uma destas escalas espaciais engloba uma série de Grupos de Trabalho, que abordam aspectos específicos dentro de cada uma destas escalas

Zona Costeira (Local)

Plataforma Continental (Regional)

Oceano (Global)

Grupos de Trabalho

Cada uma das escalas espaciais de investigação do INCT AmbTropic inclui vários Grupos de Trabalho (GTs). Estes GTs investigarão aspectos específicos dentro de cada uma destas escalas espaciais. A escolha dos temas dos Grupos de Trabalho e suas áreas específicas de investigação foram balizadas pelos seguintes principios:

(i) contemplar a heterogeneidade espacial e temporal nas suas três escalas principais de investigação, privilegiando gradientes fisico-quimicos e dinâmicos (ondas, marés, CO2, circulação etc) e ambientes marinhos intrinsecamente associados à região tropical do Brasil, tais como recifes de corais, manguezais, recursos vivos, ilhas oceânicas, processos oceânicos etc.,

(ii) construir sobre a experiência dos grupos de pesquisa locais,

(iii) concentrar os esforços dos GTs em areas geográficas comuns, não só para minimizar custos, como também estimular a sinergia entre os GTs, e

(iv) avaliação de recursos naturais (vivos e não vivos) específicos da região.

Coordenadores

Organograma

Publicações

Contribuição 1 - Veleda, D., M. Araujo, R. Zantopp, and R. Montagne (2012), Intraseasonal variability of the North Brazil Undercurrent forced by remote winds, J. Geophys. Res., 117, C11024, doi:10.1029/2012JC008392.

Intraseasonal variability of the North Brazil Undercurrent forced by remote winds

D. Veleda,1,2 M. Araujo,1,2 R. Zantopp,3 and R. Montagne4
Received 27 July 2012; revised 9 October 2012; accepted 10 October 2012; published 21 November 2012.
[1] Intraseasonal signals with periods of 2 to 3 weeks in near-surface alongshore current
measurements are detected from four moorings (K1–K4) deployed from 2000 to 2004
at the 11 S section close to the Brazilian coast as part of the German CLIVAR Tropical
Atlantic Variability Project. This section crosses the path of the North Brazil Undercurrent,
the most powerful western boundary current in the South Atlantic Ocean. We investigate the
origin of this intraseasonal variability of the North Brazil Undercurrent by relating
the oceanic oscillation of the alongshore currents to its atmospheric counterpart, the
meridional wind stress. On average, the results indicate a well-defined lagged (10 days)
correlation (0.6) structure between meridional wind stress and alongshore currents.
The oceanic region with the highest cross-correlations is identified as a relatively narrow
band along the Brazilian coast, from 22 –36 S and 40 –50 W, bounded in the north by an
eastward change in coastline orientation. The cross-wavelet transform establishes the
common power between the time series of meridional wind stress and alongshore currents,
predominantly during austral winter and spring. These signals propagate equatorward with
an alongshore speed of 285  63 km day1
, consistent with Coastal Trapped Wave theory.
Citation: Veleda, D., M. Araujo, R. Zantopp, and R. Montagne (2012), Intraseasonal variability of the North Brazil Undercurrent
forced by remote winds, J. Geophys. Res., 117, C11024, doi:10.1029/2012JC008392.

  1. Introduction
    [2] Over the past few decades, the western boundary currents of the tropical South Atlantic and their associated scales
    of variability have been investigated by numerous researchers [Molinari, 1983; Stramma, 1991; Schott and Böning,
    1991; Mayer and Weisberg, 1993; Schott et al., 1993;
    Rhein et al., 1995; Stramma et al., 1995; Dengler et al., 2004;
    Schott et al., 2005]. While the Deep Western Boundary
    Current (DWBC) transports cold North Atlantic Deep Water
    far into the southern hemisphere, the North Brazil Undercurrent (NBUC) and North Brazil Current (NBC) serve as a
    northward warm-water conduit as part of the thermohaline
    overturning cell [Gordon, 1986; Schmitz, 1995]. The South
    Atlantic Central Water (SACW), located at depths between
    100 and 500 m, is transported westward within the southern
    band of the South Equatorial Current (sSEC) until it reaches
    the Brazilian shelf. After the sSEC bifurcates, its southward
    limb becomes the Brazil Current (BC) and merges into the
    South Atlantic subtropical gyre system. The northward limb
    feeds into the NBUC [Stramma and Schott, 1999; Stramma
    et al., 2005] and - as a western boundary current - carries
    warm waters of South Atlantic origin across the equator and
    into the northern hemisphere. It also supplies the eastward
    flow of the South Equatorial Countercurrent (SECC) which
    partially recirculates into the central band of the SEC (cSEC).
    In addition, in the 50–300 m depth range, the NBUC seems to
    play an important role in the Atlantic Equatorial Gyre [Schott
    et al., 2005] and in the coupled ocean-atmosphere system.
    This strong western boundary current has a nearshore core
    position approximately 50 km from the Brazilian coast,
    reaching down to about 900 m in depth, with a maximum
    speed of about 65 cm s1 at 180 to 250 m on average. At 11 S,
    the mean flow structure of the NBUC is already well developed, indicating that the bifurcation of the sSEC is located well
    south of this section with a maximum northward NBUC flow
    in July and a minimum in the October–November period
    [Schott et al., 2005]. The sSEC bifurcation has a southernmost
    position in July and a northernmost position in November
    [Rodrigues et al., 2007; Silva et al., 2009].
    [3] While the dominant fluctuation in the NBUC core,
    approx. 50 km from the coast, has a period of about two
    months [Schott et al., 2005], biweekly signals are the dominant modes along the western (onshore) flank of the NBUC,
    about 10 km from the coast [von Schuckmann, 2006].
    [4] The cause of these intraseasonal fluctuations in the
    upper-ocean circulation may be found in several dynamic
    processes, such as local wind-forcing, remote wind-forcing via
    waveguide dynamics, mean flow instability, and resonance
    1
    Laboratório de Oceanografia Física Estuarina e Costeira, Departamento
    de Oceanografia, Universidade Federal de Pernambuco, Recife, Brazil. 2
    CEERMA-Centro de Estudos e Ensaios em Risco e Modelagem
    Ambiental, Universidade Federal de Pernambuco. 3
    GEOMAR Helmholtz Centre for Ocean Research Kiel, Kiel, Germany.
    4
    Departamento de Física, Universidade Federal Rural de Pernambuco,
    Recife, Brazil.
    Corresponding author: D. Veleda, Laboratório de Oceanografia Física
    Estuarina e Costeira, Departamento de Oceanografia, Universidade Federal
    de Pernambuco, Av. Arquitetura s/n, 50740-550, Cidade Universitária,
    Recife, PE, Brazil. (doris.veleda@ufpe.br)
    ©2012. American Geophysical Union. All Rights Reserved.
    0148-0227/12/2012JC008392
    JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 117, C11024, doi:10.1029/2012JC008392, 2012
    C11024 1 of 10
    due to the coastline geometry [von Schuckmann, 2006]. Furthermore, if these fluctuations are in reasonable proximity of
    the coast, the motion of oceanic waters over the continental
    shelf and slope may also be influenced by the earth’s rotation,
    density stratification, the offshore current regime, sloping
    bottom topography and the presence of the coastline [Allen,
    1980], causing coastal-trapped waves (hereafter CTWs). The
    existence of CTWs depends entirely on the presence of a
    shallow shelf between the coast and the deep ocean. CTWs
    propagate along the continental slopes and shelves of the
    world’s coastlines, with the coast to the right in the Northern
    Hemisphere and to the left in the Southern Hemisphere. They
    have periods of days to 2–3 weeks and wavelengths on the
    order of 2000 km, determined by the atmospheric weather
    patterns. The impact of CTWs on shelf currents and sea level
    varies with seasonal changes in stratification over the coast.
    For arbitrary topography and stratification, there is a free-wave
    mode (similar to the barotropic Kelvin wave) plus an infinite
    set of higher mode, more slowly propagating waves [Brink,
    1991]. These perturbations induce variations in sea level and
    alongshore currents over the continental shelf for typical periods ranging from the inertial period to about 20 days. Winds
    and tides in coastal regions at sub-inertial frequencies force
    these waves with periods ranging from days to weeks. Since
    the continental shelf is a transition zone between the coastal
    waters and the open ocean, the CTWs transport materials (e.g.,
    nutrients and pollutants) from regions near the coastline to and
    from the open ocean, and along the continental margins. Thus,
    CTWs modulate the upwelling system and primary productivity in the coastal regions.
    [5] This paper investigates the relationship between
    observed intraseasonal fluctuations in alongshore current
    measurements near the Brazilian shelf and the large-scale
    wind stress field over the South Atlantic through the use of
    cross-correlations and continuous wavelet transforms.
  2. Data
    2.1. Moored Array Observations at 11 S
    [6] The ocean data set used here was obtained between
    March 2000 and August 2004 as part of the German CLIVAR Tropical Atlantic Variability Project [Schott et al.,
    2005]. The data were collected in an array of five moorings
    (K1–K5) stretching 225 km across the NBUC (10 S–11.5 S)
    (Figure 1). Data sources, positions and sampling periods are
    presented in Table 1. Since the offshore mooring K5 was
    deployed for only two years, and observations were confined
    to depths below 1500 m. this mooring was not used in the
    present analysis. Additional details of this data set are given
    by Schott et al. [2005].
    Figure 1. Schematic representation of the Atlantic Subtropical Cell (STC) circulation, including subduction (blue) and upwelling (green) zones. Current branches involved in the STC flows are NEC, SEC,
    sSEC, NECC and EUC; NEUC, SEUC = North and South Equatorial Undercurrent; NBC, NBUC = North
    Brazil Current and Undercurrent; GD, AD = Guinea and Angola domes. Interior equatorward thermocline
    pathways dotted. Adapted from Schott et al. [2004]. The mooring array at 11 S is shown by the red line.
    Table 1. Location and Sampling Periods for Moorings and Wind
    Stress Data
    Data Source Latitude/Longitude Sampling Period
    Moorings, German CLIVAR Project
    Station K1 10 16.0′S
    35 51.7′W
    02/2002–01/2003
    06/2003–08/2004
    Station K2 10 22.8′S
    35 40.8′W
    02/2002–05/2004
    Station K3 10 36.7′S
    35 23.4′W
    05/2003–08/2004
    Station K4 10 56.5′S
    34 59.5′W
    01/2002–08/2004
    Wind stress data, NOAA/NCDC 40 S–0 N
    60 W–20 E
    01/2002–12/2004
    C11024 VELEDA ET AL.: ISV OF THE NBUC FORCED BY REMOTE WINDS C11024
    2 of 10
    2.2. Wind Stress Data
    [7] The wind stress data set used in this study was obtained
    from NOAA/NCDC (National Oceanic and Atmospheric
    Administration/National Climatic Data Center) and is available
    at www.ncdc.noaa.gov/oa/rsad/seawinds.html. The domain
    used here covers the area 40 S to 0 N and 60 W to 20 W
    (Figure 2). In this database, surface wind stresses (N m2
    ) are
    estimated from blended sea surface wind speeds at 10 m above
    sea level, generated from six satellites, on a global 0.25 regular
    grid and for several time resolutions [Zhang et al., 2006]. Wind
    speeds are converted to (u, v) components using the NCEP
    Reanalysis 2 (NRA-2). In this work, the wind stress data has a
    temporal resolution of 12 h.
    [8] The mean zonal and meridional components of the
    wind stress are shown in Figures 2a–2d for austral summer
    (DJF) and winter (JJA). These charts capture the main spatial
    structure of the observed means, in agreement with the published literature [Trenberth et al., 1990; Harrison, 1989;
    Castelão and Barth, 2006]. During austral summer, the zonal
    wind is negative (easterlies) over nearly the entire domain,
    while positive (northward) meridional wind stress is
    restricted to equatorial latitudes in the western part of the
    South Atlantic Ocean and negative (southward) for almost
    the entire western South Atlantic (Figures 2a and 2b). The
    zonal components of the wind stress are strongest in June,
    July and August. Northward meridional wind stress is
    observed in the South Atlantic during JJA, expanding equatorward and reaching 20 S in the western basin (Figures 2c
    and 2d). Negative (southward) meridional wind stress is
    found between 20 and 30 S, and positive (northward) to the
    south of it.
  3. Results
    3.1. Current Structure at 11 S
    [9] The current meter time series were smoothed using a
    40-h lowpass filter to eliminate tidal and inertial effects.
    Current vectors were divided into alongshore and cross-shore
    components by rotating clockwise by 36 , i.e., parallel to the
    coastline. Figure 3 shows the mean section of alongshore
    velocities at 11 S for the entire deployment length (2000–
    2004); instrument locations for moorings K1–K4 are marked
    by black dots. The strongest contribution to the NBUC
    transports occurs near 250 m depth at moorings K1 and K2
    [Schott et al., 2002; Stramma et al., 2003; Schott et al.,
    2005], indicated by the strong subsurface core in the upper
    left corner of Figure 3. Below the NBUC, we find the colder
    water masses of the DWBC, with the core centered at 1500–
    3500 m depth along mooring K3 (dark blue colors). Mooring
    K4 is located outside the equatorward boundary current, with
    a weak mean southward flow between 500 and 3500 m depth.
    [10] Fluctuations about these means are shown as anomalies of the 100 m current vectors (i.e., demeaned), as at this
    level the measurements are available for all moorings K1–K4
    (Figure 4). The dominant intraseasonal signals have periodicities of 2–3 weeks and 2–3 months (see also Figure 5).
    The highest amplitudes at periods of 2–3 weeks are found in
    the coastal boundary region at moorings K1 and K2, gradually decreasing away from the NBUC core. Figure 5 shows
    Figure 2. (a, c) Mean zonal and (b, d) mean meridional components of wind stress for austral summer
    (DJF) and winter (JJA). The zero contours are represented by dashed lines.
    C11024 VELEDA ET AL.: ISV OF THE NBUC FORCED BY REMOTE WINDS C11024
    3 of 10
    Figure 4. Vector time series of 40-h low-pass filtered alongshore current anomalies at moorings K1–K4
    at 100 m depth, from March 2000 to August 2004.
    Figure 3. Mean of alongshore velocity at 11 S, derived from moorings K1–K4. Instrument locations are
    marked by black dots. Note the equatorward NBUC above 1000 m and the southward flowing DWBC
    between 1500 and 3500 m (adapted from von Schuckmann [2006]).
    C11024 VELEDA ET AL.: ISV OF THE NBUC FORCED BY REMOTE WINDS C11024
    4 of 10
    the spectral distribution of currents between 50 and 300 m
    depth for moorings K1 to K4. High-frequency fluctuations
    (10–30 days) are more prominent near the coast, at mooring
    K1. Farther offshore, at moorings K2 to K4, these fluctuations have the highest energy in the upper layers and scale
    down with increasing distance from the NBUC current core.
    Note that the variance levels at mooring K2 are nearly identical between 150 and 300 m.
    3.2. Cross-Correlation Between Wind Stress
    and Currents
    [11] We used cross-correlation analysis between the
    meridional wind stress and alongshore currents to spatially
    localize the remote forcing across the entire ocean basin.
    Previous work indicates that the meridional component of
    wind stress is the prime generator of such coastal-trapped
    waves [Brink, 1991; Csanady, 1997], and the periods
    involved correspond to those of the weather systems. The
    meridional wind stress components are cross-correlated with
    alongshore currents at 100 m depth at 11 S for moorings
    K1–K4 (Figure 6) as this depth level is available for all
    moorings (see Figure 5).
    [12] For mooring K1, two separate intervals are shown,
    caused by a gap in the available current measurements. Since
    we are interested in the intraseasonal components of the signal
    in the 10–30 day band, we applied a 10–30 day band-pass
    filter to both the atmospheric and oceanic data. The correlation
    of the meridional wind stress for each grid point with the K1
    alongshore current at 100 m depth is mapped for the entire
    ocean basin (Figure 6a), and the lag between the two signals is
    shown in the maps of the right panels. Identical procedures
    were used for the other moorings, displayed in Figures 6b–6e.
    Clearly, the lag between the wind stress of a given grid point
    and the 11 S currents decreases with decreasing distance to
    the moorings, showing a systematic propagation along the
    shore. For the first interval of available data at mooring K1,
    the maximum cross-correlation between meridional wind
    and alongshore current is found between 22 and 36 S, which
    is approximately 1500–2000 km from the 11 S moorings
    (Figure 6a). The corresponding lag is 6.5 days. For the
    2003–2004 period of mooring K1 (Figure 6b), there are
    Figure 5. Variance preserving spectra of kinetic energy [cm2 s
    2
    ] (solid line) for moorings K1–K4, at 50
    to 300 m depth, from March 2000 to August 2004, with the respective 95% confidence limits.
    C11024 VELEDA ET AL.: ISV OF THE NBUC FORCED BY REMOTE WINDS C11024
    5 of 10
    weaker cross-correlations, with alongshore currents lagging
    the wind stress by 10 days. Subsequently, we find a lag of
    9 days at mooring K2 (Figure 6c), a lag of 7 days at
    mooring K3 (Figure 6d), and a lag of 9.5 days for mooring
    K4. Here is also where we find the highest cross-correlation
    values between meridional wind stress and the alongshore
    current at about 36 S (Figure 6e).
    [13] Table 2 summarizes the cross-correlation results of
    Figure 6 between the meridional wind stress and alongshore
    current at 11 S. The coordinates of cross-correlation maxima, the distances to the mooring sites, and the lag period
    clearly indicate a strong connection between the alongshore
    current signal at 11 S and its remote atmospheric forcing
    located at about 22 S to 36 S (box area in Figure 6).
    [14] To further investigate this connection, we spatially
    averaged the meridional wind stress components across the
    area defined by the highest cross-correlations shown in
    Figure 6. The resulting time series were cross-correlated
    with the corresponding time series of meridional currents at
    11 S (moorings K1–K4) at different depths, 50 to 1000 m
    (Figure 7). The highest correlations are found in the nearsurface layers of the NBUC core, gradually weakening
    downward and outward toward the open ocean. The right
    panel shows the corresponding lags, where currents lag the
    wind-forcing by 8 to 10 days (Figure 7, right).
    3.3. Time-Frequency Common Power Between Wind
    Stress and Currents
    [15] Most traditional mathematical methods to examine
    periodicities in frequency space, such as the Fourier analysis, implicitly assume that the underlying processes are stationary in time. However, wavelet transforms can be used
    to analyze a non-stationary time series at many different
    frequencies and scales by expanding the time series into
    time-frequency space in order to find localized intermittent
    periodicities. The wavelet transform is based on a practical
    guide to wavelet analysis of Torrence and Compo [1998].
    One of the problems with the resulting “global wavelet
    spectra” is a distortion of the energy distribution [Liu et al.,
    2007; Veleda et al., 2012], with high-frequency peaks
    shown lower than their low-frequency counterparts. In this
    work, we construct rectified Cross Wavelet Transform
    (XWT) by normalizing the wavelet transform by the square
    root of the scale, based on Veleda et al. [2012].
    [16] An XWT analysis was applied to meridional wind
    stress and alongshore currents at 100 m depth (moorings
    K1–K4). The XWT identifies the regions in time-frequency
    space with a large common power between two time series.
    The individual cross-wavelet spectra, from 2002 to 2004, are
    averaged to form a seasonal ensemble cross-wavelet spectrum, for austral winter (June to November) and austral
    summer (December to May), respectively. In this case, we
    used the wind stress data restricted to the region where the
    highest cross-correlations between the wind stress and
    alongshore currents were found (see Figure 6).
    Figure 6. (left) The cross-correlation between meridional
    wind stress and alongshore currents at 100 m depth for
    (a) K1 mooring, 3/2002–1/2003 and (b) 6/2003–8/2004,
    © K2 mooring, 3/2002–5/2004, (d) K3 mooring, 6/2003–
    8/2004 and (e) K4 mooring, 3/2002–5/2004. (right) The
    corresponding lags (currents lagging winds) for the crosscorrelation. Box in cross-correlation maps marks the area
    of maximum cross-correlation.
    Table 2. Coordinates of Maximum Cross-Correlation Between Meridional Wind Stress and Alongshore Currents, Corresponding
    Distances and Lag Periodsa
    Mooring Maximum Cross-Correlation Position Distance (km) Lag Period (days)
    K1_I 0.55 (0.076) 25 S–44 W 1550 6.5
    K1_II 0.44 (0.067) 34 S–47 W 2550 10
    K2 0.52(0.049) 28 S–45 W 1890 8.5
    K3 0.64 (0.065) 33 S–48 W 2440 7
    K4 0.39 (0.046) 36 S–50 W 2780 9.5
    a
    The 95% confidence limits are shown in parentheses.
    C11024 VELEDA ET AL.: ISV OF THE NBUC FORCED BY REMOTE WINDS C11024
    6 of 10
    [17] The seasonal ensemble averages of the XWT between
    the meridional wind stress averaged over the high crosscorrelation area [22 –36 S] and the corresponding alongshore currents within the matching time interval are shown
    in Figure 8, with the blocked out “white” margins indicating
    the “Cone of Influence” (COI), where edge artifacts become
    important [Grinsted et al., 2004]. The color scale for
    Figure 8 shows maxima in dark red and minima in dark blue.
    [18] Figure 8 clearly indicates that the XWT between currents and wind stress is strongest for the 16-day period during
    the month of September, linking this periodicity band of
    the NBUC, at 11 S for mooring K1 (winter) at 100 m depth,
    to the wind stress forcing at 22 –36 S close to the Brazilian
    coast. The ensemble average XWT for the winter season
    shows a stronger or more prominent response of the 16-day
    period of the alongshore currents to the wind stress.
    [19] An identical analysis was done for moorings K2, K3
    and K4 (Figure 8). The result shows that during austral
    winter, the 16-day periodicity of the 100 m currents at
    moorings K1–K4, along the western and eastern flank of the
    NBUC at 11 S, is linked to the wind-forcing at 22 –36 S
    along the coast of Brazil. At mooring K1, the highest energy
    spans the period from July through October, except for
    August. For mooring K2, the highest energy is centered in
    July and weaker in September and November. For moorings
    K3 and K4, the highest energy extents from July through
    October.
  4. Discussion
    [20] Previous studies in other regions have confirmed the
    presence of CTWs with typical periods of 5–20 days. Brink
    [1983] associated these waves with remote wind-forcing,
    while Battisti and Hickey [1984] determined that the pressure and alongshore velocity fields in the Northwest Pacific
    may be a response of the wind-forced CTW. The variance in
    sea surface pressure off Oregon and Washington was found
    to be generated by wind forcing between San Francisco and
    Cape Mendocino, California. The signal propagates 900 to
    1300 km northward to Washington-Oregon, arriving there
    3 to 4 days later. Brink [1982] also pointed out the agreement
    between CTW theory and observations off Peru in 1977,
    indicating free wave phase speeds of about 200 km day1
    ,
    as well as sea level and alongshore velocity fluctuations in
    the 5–10 day period band. Spillane et al. [1987] found
    oscillations with intraseasonal periods of 36–73 days close to
    the coast of Peru, with poleward phase propagation of 150–
    200 km day1
    . Enfield [1987] established that the intraseasonal sea level variations pointed out by Spillane et al. [1987]
    are forced in the western equatorial Pacific by atmospheric
    oscillations, consistent with previous studies with propagation speeds of 216–259 km day1
    .
    [21] Smith [1978] found a persistent poleward propagation
    of fluctuations in currents and sea level along the Peruvian
    coast between 10 and 15 S, with a wave speed of about
    200 km day1
    . Enfield and Allen [1980] also estimated the
    wave phase speed as the ratio of alongshore station separation to the corresponding lag. Along the Peruvian coast,
    Camayo and Campos [2006] used wavelet analysis to study
    the intraseasonal current oscillations, suggesting remotely
    forced baroclinic Kelvin waves with periods between 10 and
    20 days, with velocities of approx. 200 km day1
    .
    [22] In this work, we establish the propagation of Coastal
    Trapped Waves along the Brazilian coast as a suitable
    mechanism for explaining the strong correlation between the
    meridional wind stress at latitudes of 22 S–36 S and the
    Figure 7. Cross-correlation and lag (in days) between meridional wind stress averaged in the area along
    the Brazilian coast [22 S–36 S]–[40 W–50 W] and alongshore currents at the 11 S section.
    C11024 VELEDA ET AL.: ISV OF THE NBUC FORCED BY REMOTE WINDS C11024
    7 of 10
    alongshore currents at 11 S. Kelvin waves are low-frequency
    gravity waves which occur where the deflection caused by
    the Coriolis force is either constrained by coastlines or goes
    to zero at the equator. These waves have typical amplitudes
    of several tens of meters in the thermocline region, and
    lengths are thousands of kilometers.
    [23] Spectral analysis of the current velocities identified
    high-frequency components present near the surface and
    gradually decreasing with depth. A fluctuation at 10–30 day
    periodicity is observed in the four moorings at the jointly
    available 100 m depth level. At 11 S, the thermocline is located
    at about 65 m depth [Schott et al., 2005; von Schuckmann,
    2006], so that the core of the NBUC is located just below
    these strong density and temperature gradients, making this
    flow regime potentially favorable for internal Kelvin wave
    propagation.
    [24] Internal coastal-trapped Kelvin waves may be generated by an abrupt change in the winds and depend on the
    existence of a coast against which they can lean. Our work
    showed maximum cross-correlations for the nearshore wind
    data located between 22 S and 36 S, a region along the
    Brazilian coast usually referred to in the literature as the
    Southeast Brazil Bight (SBB) [Castro Filho, 1985; Stech and
    Lorenzzetti, 1992; Campos et al., 1995; Cirano and Campos,
    1996]. The northern boundary of this area corresponds to a
    relatively narrow shelf close to Cabo Frio (22 S), where an
    abrupt change in coastline orientation exists. Furthermore,
    the SBB is also under the influence of synoptic and
    mesoscale frontal low pressure systems which interact
    with the western reaches of the South Atlantic High
    (or “St. Helena High”). Strong meridional winds from the
    northeast (NE) occur before the passage of a low pressure
    system, then rapidly changing to southwesterly winds (SW).
    These frontal systems have frequencies of 3 to 4 per
    month [Rodrigues et al., 2004], with average speeds of
    500 km day1
    , thereby crossing the SBB region in about
    2 days [Stech and Lorenzzetti, 1992]. As a result, these atmospheric anomalies induce drastic changes in wind direction
    and cause significant disturbances in the ocean, such as mean
    sea level changes, generation of surface waves and currents
    [Castro Filho, 1985; Campos et al., 1995]. We argue that the
    combination of local wind stress variability and the abrupt
    change in coastline direction reveals the SBB as a preferred
    locus for CTW generation along the Brazilian shore. The
    distance between the currents at 11 S and the area where
    high cross-correlations were found is about 1600–2700 km.
    In addition, the ratio between these distances and their
    corresponding lag periods yields an equatorward propagation
    speed of 285  63 km day1 along the Brazilian shore.
    Coastal trapped waves certainly have major effects on sea
    level and currents locally, but their far-reaching consequences
    may not yet be fully explored: As the waves approach the
    equator, they continue eastward as equatorial Kelvin waves,
    only to be dispersed again as coastal Kelvin waves along the
    eastern boundary (off Africa) and also reflected westward as
    equatorial Rossby waves. The conclusions regarding our
    Figure 8. The seasonal ensemble average of the Cross-Wavelet Transform between the meridional wind
    stress averaged in the area along the Brazilian coast [22 S–36 S]–[40 W–50 W] and the respective alongshore current.
    C11024 VELEDA ET AL.: ISV OF THE NBUC FORCED BY REMOTE WINDS C11024
    8 of 10
    observations may have significant effects on the entire equatorial waveguide system.
  5. Summary
    [25] The western boundary regime of the tropical South
    Atlantic is well known for its complex dynamics and multiple
    scales of variability, with the North Brazil Undercurrent
    (NBUC) playing an important role in the Atlantic Subtropical
    Cell (STC) and in the coupled ocean-atmosphere system.
    [26] Previous studies of these moorings (K1–K4) at 11 S
    section have shown that the NBUC core, about 50 km offshore, has a dominant fluctuation with a periodicity of two
    months [Schott et al., 2005; von Schuckmann, 2006]. In this
    work we investigate the forcing mechanism behind additional
    intraseasonal periodicities of the near-surface alongshore current measurements at 2 to 3 weeks. Cross-correlation analysis
    between currents and meridional wind stress points to the wind
    field located between 22 and 36 S close to Brazilian coast as
    the driving mechanism.
    [27] The distances between the wind-forcing area and the
    mooring locations, as well as the corresponding lags, indicate
    an equatorward propagation speed of 285  63 km day1
    along the Brazilian coast. Such distances and travel times
    suggest Coastal Trapped Waves (CTWs) as the prime candidates for accomplishing this task. Therefore, two of the
    main ingredients for the existence of CTW are readily
    available, namely wind stress variability and abrupt change
    in coastline direction.
    [28] We investigated the relationship between the NBUC,
    wind stress and CTWs by using several statistical and mathematical tools. High-frequency variability is the dominant
    mode along the western flank of the NBUC. Spectral analysis
    of the current velocities for K1 (the mooring closest to the
    coast) shows a spectral peak at 14 and 30 days at 100 m depth.
    This peak decreases as we move farther offshore. Crosscorrelation analysis of currents and wind stress along the entire
    Brazilian coast reveals the precise origin of these fluctuations.
    In addition, the Cross-Correlation Wavelet Transform accurately identifies the main structures in the frequency-time
    space where the atmospheric and oceanic time series both had
    their largest common power.
    [29] The correlations between meridional wind stress, from
    22 to 36 S near the Brazilian coast, and alongshore currents
    at 11 S are strongest during austral winter and spring. As
    shown in Enfield and Allen [1980], these signals propagate
    equatorward with the coast to the left (in the southern hemisphere) and have alongshore speeds consistent with wave
    propagation processes. Furthermore, the area of origin for the
    forcing of intraseasonal current signals is dominated by the
    first baroclinic mode over the continental slope. It has also
    been shown that this area has stronger density stratification
    during austral winter than austral summer [Campos et al.,
    2000], making it more favorable for the propagation of
    internal Kelvin waves.
    [30] Acknowledgments. The authors would like to thank the scientific team of the German Climate Variability and Predictability (CLIVAR)
    program and, in particular, our colleagues from GEOMAR for providing
    full access to the K1–K4 mooring data. We also thank the editor and the
    two anonymous reviewers for their valuable suggestions and comments.
    The first author thanks Peter Brandt and Lothar Stramma for their assistance
    during her one-year stay at GEOMAR in Kiel. She also wishes to thank
    CAPES/DAAD (Coordination for the Improvement of Higher Education
    Staff/Deutscher Akademischer Austauschdienst) for scholarship support.
    R.M. acknowledges financial support from the Pernambuco State Agency
    FACEPE (APQ-0871-1.05/08).This work was carried out under the National
    Institute on Science and Technology in Tropical Marine Environments -
    INCT-AmbTropic (CNPq process 565054/2010-4), and under the project
    BIO-NE (CNPq process 558143/2009-1.
    References
    Allen, J. S. (1980), Models of wind-driven currents on the continental
    shelf, Annu. Rev. Fluid Mech., 12, 389–433, doi:10.1146/annurev.
    fl.12.010180.002133.
    Battisti, D. S., and B. M. Hickey (1984), Application of remote windforced coastal trapped wave theory to the Oregon and Washington coasts,
    J. Phys. Oceanogr., 14, 887–903, doi:10.1175/1520-0485(1984)014<0887:
    AORWFC>2.0.CO;2.
    Brink, K. H. (1982), A comparison of long coastal trapped wave theory with
    observations off Peru, J. Phys. Oceanogr., 12, 897–913, doi:10.1175/
    1520-0485(1982)012<0897:ACOLCT>2.0.CO;2.
    Brink, K. H. (1983), The near-surface dynamics of coastal upwelling, Prog.
    Oceanogr., 12, 223–257, doi:10.1016/0079-6611(83)90009-5.
    Brink, K. H. (1991), Coastal-trapped waves and wind-driven currents over
    the continental shelf, Annu. Rev. Fluid Mech., 23, 389–412, doi:10.1146/
    annurev.fl.23.010191.002133.
    Camayo, R., and E. J. D. Campos (2006), Application of wavelet transform
    in the study of coastal trapped waves off the west coast of South America,
    Geophys. Res. Lett., 33, L22601, doi:10.1029/2006GL026395.
    Campos, E. J. D., J. E. Gonçalves, and Y. Ikeda (1995), Water mass characteristics and geostrophic circulation in the South Brazil Bight: Summer
    of 1991, J. Geophys. Res., 100(C9), 18,537–18,550, doi:10.1029/
    95JC01724.
    Campos, E. J. D., D. Velhote, and I. C. A. da Silveira (2000), Shelf break
    upwelling driven by Brazil current cyclonic meanders, Geophys. Res.
    Lett., 27, 751–754, doi:10.1029/1999GL010502.
    Castelão, R. M., and J. A. Barth (2006), Upwelling around Cabo Frio, Brazil:
    The importance of wind stress curl, Geophys. Res. Lett., 33, L03602,
    doi:10.1029/2005GL025182.
    Castro Filho, B. M. (1985), Subtidal response to wind forcing in the South
    Brazil Bight during winter, PhD thesis, 211 pp., Univ. of Miami, Miami,
    Fla.
    Cirano, M., and E. J. D. Campos (1996), Numerical diagnostic of the circulation in the Santos Bight with COROAS hydrographic data, Braz.
    J. Oceanogr., 44(2), 105–121.
    Csanady, G. T. (1997), On the theories that underlie our understanding of
    continental shelf circulation, J. Oceanogr., 53, 207–229.
    Dengler, M., F. A. Schott, C. Eden, P. Brandt, J. Fischer, and R. Zantopp
    (2004), Break-up of the Atlantic deep western boundary current into
    eddies at 8 S, Nature, 432, 1018–1020, doi:10.1038/nature03134.
    Enfield, D. B. (1987), The intraseasonal oscillation in eastern Pacific sea
    levels: How is it forced?, J. Phys. Oceanogr., 17, 1860–1876,
    doi:10.1175/1520-0485(1987)017<1860:TIOIEP>2.0.CO;2.
    Enfield, D. B., and J. S. Allen (1980), On the structure and dynamics of
    monthly mean sea level anomalies along the Pacific coast of North and
    South America, J. Phys. Oceanogr., 10, 557–578, doi:10.1175/1520-
    0485(1980)010<0557:OTSADO>2.0.CO;2.
    Gordon, A. L. (1986), Interocean exchange of thermocline water, J. Geophys.
    Res., 91, 5037–5046, doi:10.1029/JC091iC04p05037.
    Grinsted, A., J. C. Moore, and S. Jevrejeva (2004), Application of the crosswavelet transform and wavelet coherence to geophysical time series,
    Nonlinear Processes Geophys., 11, 561–566, doi:10.5194/npg-11-561-

Harrison, D. E. (1989), On climatological monthly mean wind stress and
wind stress curl fields over the world ocean, J. Clim., 2, 57–70,
doi:10.1175/1520-0442(1989)002<0057:OCMMWS>2.0.CO;2.
Liu, Y., X. S. Liang, and R. H. Weisberg (2007), Rectification of the bias in
the wavelet power spectrum, J. Atmos. Oceanic Technol., 24(12), 2093–
2102, doi:10.1175/2007JTECHO511.1.
Mayer, D. A., and R. H. Weisberg (1993), A description of COADS
surface meteorological fields and the implied Sverdrup transports for the
Atlantic Ocean from 30 S to 60 N, J. Phys. Oceanogr., 23, 2201–2221,
doi:10.1175/1520-0485(1993)023<2201:ADOCSM>2.0.CO;2.
Molinari, R. L. (1983), Observations of near-surface currents and temperature in the central and western tropical Atlantic Ocean, J. Geophys. Res.,
88, 4433–4438.
Rhein, M., L. Stramma, and U. Send (1995), The Atlantic Deep Western
Boundary Current: Water masses and transports near the equator,
J. Geophys. Res., 100, 2441–2451, doi:10.1029/94JC02355.
Rodrigues, M. L. G., D. Franco, and S. Sugahara (2004), Climatologia de
frentes frias no litoral de Santa Catarina, Braz. J. Geophys., 22(2), 135–151.
C11024 VELEDA ET AL.: ISV OF THE NBUC FORCED BY REMOTE WINDS C11024
9 of 10
Rodrigues, R. R., L. M. Rothstein, and M. Wimbush (2007), Seasonal variability of the South Equatorial Current bifurcation in the Atlantic Ocean:
A numerical study, J. Phys. Oceanogr., 37, 16–30, doi:10.1175/
JPO2983.1.
Schmitz, W. J., Jr. (1995), On the interbasin-scale thermohaline circulation,
Rev. Geophys., 33, 151–173, doi:10.1029/95RG00879.
Schott, F., and C. W. Böning (1991), The WOCE model in the western
equatorial Atlantic: Upper-layer circulation, J. Geophys. Res., 96,
6993–7004, doi:10.1029/90JC02683.
Schott, F., J. Fischer, J. Reppin, and U. Send (1993), On mean and seasonal
currents and transports at the western boundary of the equatorial Atlantic,
J. Geophys. Res., 98, 14,353–14,368, doi:10.1029/93JC01287.
Schott, F., P. Brandt, M. Hamann, J. Fisher, and L. Stramma (2002),
On the boundary flow off Brazil at 5–10 S and its connection to the interior tropical Atlantic, Geophys. Res. Lett., 29(17), 1840, doi:10.1029/
2002GL014786.
Schott, F. A., J. P. McCreary, and G. C. Johnson (2004), Shallow overturning circulation of the tropical-subtropical oceans, in Earth’s Climate: The
Ocean Atmosphere Interaction, Geophys. Monogr. Ser., vol. 147, edited
by C. Wang, S.-P. Xie, and J. A. Carton, pp. 261–304, AGU, Washington,
D. C., doi:10.1029/147GM15
Schott, F., M. Dengler, R. Zantopp, L. Stramma, J. Fischer, and P. Brandt
(2005), The shallow and deep western boundary circulation of the South
Atlantic at 5 –11 S, J. Am. Meteorol. Soc., 35, 2031–2053.
Silva, M. A., M. Araújo, J. Servain, P. Penven, and C. Lentini (2009), Highresolution regional ocean dynamics simulation in the southwestern tropical
Atlantic, Ocean Modell., 30, 1–29, doi:10.1016/j.ocemod.2009.07.002.
Smith, R. L. (1978), Poleward propagating perturbations in currents and
sea levels along the Peru coast, J. Geophys. Res., 83, 6083–6092,
doi:10.1029/JC083iC12p06083.
Spillane, M. C., D. B. Enfield, and J. S. Allen (1987), Intraseasonal oscillations in sea level along the west coast of the Americas, J. Am. Meteorol.
Soc., 17, 313–325.
Stech, J. L., and J. A. Lorenzzetti (1992), The response of the South Brazil
Bight to the passage of wintertime cold fronts, J. Geophys. Res., 97(C6),
9507–9520, doi:10.1029/92JC00486.
Stramma, L. (1991), Geostrophic transport of the South Equatorial
Current in the Atlantic, J. Mar. Res., 49, 281–294, doi:10.1357/
002224091784995864.
Stramma, L., and F. Schott (1999), The mean flow field of the tropical
Atlantic Ocean, Deep Sea Res., Part II, 46, 279–303, doi:10.1016/
S0967-0645(98)00109-X.
Stramma, L., J. Fischer, and J. Reppin (1995), The North Brazil Undercurrent, Deep Sea Res., Part I, 42(5), 773–795, doi:10.1016/0967-0637(95)
00014-W.
Stramma, L., J. Fischer, P. Brandt, and F. Schott (2003), Circulation, variability and near-equatorial meridional flow in the central tropical Atlantic,
in Interhemispheric Water Exchanges in the Atlantic Ocean, edited by
G. J. Goni and P. Malanotte-Rizzoli, pp. 1–22, Elsevier, Amsterdam.
Stramma, L., M. Rhein, P. Brandt, M. Dengler, C. Böning, and M. Walter
(2005), Upper ocean circulation in the western tropical Atlantic in boreal
fall 2000, Deep Sea Res., Part I, 52, 221–240, doi:10.1016/j.dsr.2004.
07.021.
Torrence, C., and G. P. Compo (1998), A practical guide to wavelet analysis, Bull. Am. Meteorol. Soc., 79(1), 61–78, doi:10.1175/1520-0477
(1998)079<0061:APGTWA>2.0.CO;2.
Trenberth, K. E., W. G. Large, and J. G. Olson (1990), The mean annual
cycle in global ocean wind stress, J. Phys. Oceanogr., 20, 1742–1760,
doi:10.1175/1520-0485(1990)020<1742:TMACIG>2.0.CO;2.
Veleda, D., R., Montagne, and M. Araújo (2012), Cross-wavelet bias
corrected by normalizing scales, J. Atmos. Oceanic Technol., 29,
1401–1408, doi:10.1175/JTECH-D-11-00140.1.
von Schuckmann, K. (2006), Intraseasonal variability in the southwestern
and central tropical Atlantic Ocean, PhD thesis, 140 pp., Christian
Albrechts Univ. Kiel, Kiel, Germany.
Zhang, H.-M., J. J. Bates, and R. W. Reynolds (2006), Assessment of composite global sampling: Sea surface wind speed, Geophys. Res. Lett., 33,
L17714, doi:10.1029/2006GL027086.
C11024 VELEDA ET AL.: ISV OF THE NBUC FORCED BY REMOTE WINDS C11024
10 of 10