The dispersal of the Amazon and Orinoco River water in the tropical Atlantic and Caribbean Sea: Observation from space and S-PALACE floats

https://doi.org/10.1016/j.dsr2.2004.04.001Get rights and content

Abstract

The temporal evolution of spatial patterns of the colored water mass associated with the discharges of the Amazon and Orinoco Rivers between 1997 and 2002 was examined using concurrent in situ and satellite observations in the region bounded by 0°N–24°N and 70°W–40°W. Patches of low-salinity (∼32–34) surface waters were frequently observed with Salinity Profiling Autonomous LAgrangian Current Explorer (S-PALACE) floats as far as 2000 km away from the mouths of the Amazon and Orinoco Rivers. Ocean-color data collected with the Sea-viewing Wide Field-of-view Sensor (SeaWiFS) show that these patches originate with these rivers. Chlorophyll (Chl) and colored dissolved organic matter (CDOM) distribution estimated with SeaWiFS data are well correlated with low sea-surface salinity (SSS) throughout the observation period (average correlation coefficients for monthly observations of −0.6 for the SSS–CDOM relationship and −0.5 for the SSS–Chl relationship). The color of the water in the river plumes is dominated by CDOM rather than Chl light absorption, in contrast to adjacent open-ocean and equatorial upwelling regions, where optical properties are primarily controlled by Chl. The mean CDOM absorption coefficient for the area covered by the Amazon's plume showed a peak in July for the 1998–2001 period, which lagged the Amazon's hydrograph by one month. The timing of the CDOM peak along different sections of the Amazon River plume implies an estimated advection speed of ∼35 cm s−1 within the plume. Anti-cyclonic rings (eddies) with diameters of 400–500 km were observed moving along the northeastern coast of South America every year and in all seasons, at a frequency of about one every 30–60 days. These eddies moved to the northwest at a speed of 8–20 cm s−1, with higher speeds observed during the first half of each calendar year. The average depth of the Amazon plume for SSS ranging from 32 to 35 was 20–30 m, and the total fresh-water volume contained in the plume was consistent with discharge rates. Due to uncertainties in the Chl and CDOM absorption coefficient estimates in river plumes, accurate estimation of the rivers’ impact on regional primary production remains a challenge. The paired SeaWiFS and S-PALACE float observations validate the hypothesis that the coherent surface color patches seen annually far offshore in the tropical North Atlantic are caused by dispersal of Amazon and Orinoco River water. In contrast, wind-driven upwelling near the edge of the North Brazil Current retroflection, another mechanism causing color enhancement, is not suggested by these data sets.

Introduction

River discharge plays an important role in the hydrological cycle and thermodynamic stability of the ocean. Knowledge of the variations in the extent and dispersal patterns of river plumes, and their mixing rates with oceanic water is therefore critical in all aspects of continental shelf and regional oceanography (e.g., Lohrenz et al., 1990). Yet there are few observations to help understand the fate of water derived from the world's major rivers. Of particular interest is the effect of rivers like the Amazon and the Orinoco on the adjacent ocean. The combined discharge from these rivers represents nearly 20% of the total global annual fresh-water discharge via rivers, which is all delivered into the western tropical North Atlantic. This study uses ocean-color satellite imagery and concurrent in situ hydrographic observations to examine the dispersal pattern of the Amazon and the Orinoco River waters over a period of five years.

The Amazon is the world's largest river in terms of fresh-water discharge (Milliman and Meade, 1983; Perry et al., 1996). The annual mean discharge rate onto the north Brazilian shelf at the equator is (1.93±0.13)×105 m3 s−1 (Perry et al., 1996). The Amazon's plume spreads offshore and northwestward along the north Brazilian coast, covering most of the continental shelf from 1°S to 5°N (Muller-Karger et al., 1988, Muller-Karger et al., 1995). Beyond this region, the Amazon's water has been traced northwestward into the Caribbean Sea and eastward in the North Atlantic (Muller-Karger et al., 1988, Muller-Karger et al., 1995; Johns et al., 1990).

Hydrographic surveys by Lentz and Limeburner (1995) revealed that the Amazon Plume over the shelf is typically 3–10 m thick and between 80 and >200 km wide. Both chlorophyll (Chl) concentration and primary productivity on the continental shelf are greatest in the river–ocean transition zone, where the bulk of heavy sediments settle out of surface waters (Smith and Demaster, 1996). The combination of riverine nutrient input and increased irradiance availability creates a highly productive transition zone, the location of which varies with the discharge from the river. Phytoplankton biomass and productivity of over 25 mg Chl-a m−3 and 8 g C m−2 d−1, respectively, are found in this transition region (Smith and Demaster, 1996). Because of this, the North Brazil shelf acts as a significant sink for atmospheric CO2 (Ternon et al., 2000).

Muller-Karger et al., 1988, Muller-Karger et al., 1995 discovered what appeared to be regular seasonal changes in the patterns of dispersal of the Amazon water in the western tropical Atlantic. They proposed that during the first half of the year the Amazon water generally dispersed to the northwest, toward the Caribbean Sea and over a broad geographic area. This included a coastal turbid zone, a large river plume and offshore lenses of low-salinity water. However, during the second half of the year, the Amazon plume flows around the North Brazil Current (NBC) retroflection near 5–10°N, and is carried eastward in the meandering north equatorial counter current (NECC). Johns et al. (1990) reported that the NBC retroflection regularly shed large anti-cyclonic eddies. Richardson et al. (1994) and Fratantoni and Glickson (2002) further attempted to describe the eddies and quantify the frequency at which they are generated. Muller-Karger et al., 1988, Muller-Karger et al., 1995 speculated that the river-derived water probably contained both colored dissolved organic matter (CDOM) and phytoplankton growing on river-derived nutrients, and that both contributed to the color that traced the river plume in ocean-color imagery.

Longhurst (1993) argued that the colored water seen in the NBC retroflection and NECC was due to phytoplankton blooming by action of eddy upwelling (e.g., Woods, 1988), and not due to the dispersal of Amazon water. In subsequent companion papers, Longhurst (1995) and Muller-Karger et al. (1995) further explored the two separate mechanisms proposed to explain the seasonal formation of the plume. However, both concluded that without extensive in situ data and without better imagery than that provided by the coastal zone color scanner (CZCS) (the only ocean-color data available to that date), it would be difficult to conclusively prove which of the mechanisms dominated the surface color signal observed.

Remote sensing is the only viable approach to measure the temporal variability of the dispersal of waters from large rivers such as the Amazon over synoptic scales. As we enter the era of global observing systems, we now have superior ocean-color satellite sensors as well as devices deployed in situ. We are now in an excellent position to revisit the origin of the color observed in the tropical North Atlantic with a multi-year time series of excellent remote sensing and concurrent in situ observations.

Ranked generally as the 4th of the world's rivers in terms of flow, and the largest river in Venezuela, the Orinoco discharges a mean of (3.10±0.38)×104 m3 s−1 fresh water into the southern Caribbean Sea. Discharge ranges from 1×104 m3 s−1 in March to 7×104 m3 s−1 in August (Perry et al., 1996; Muller-Karger and Aparicio Castro, 1994). Previous studies have shown that the Orinoco River plume covers the eastern and northern Caribbean Sea on a seasonal basis (Muller-Karger et al., 1989; Bidigare et al., 1993; DelCastillo et al., 1999). Using CZCS images, Muller-Karger et al. (1989) concluded that river discharge was one of the major factors controlling the pigment distribution in the wider Caribbean Sea and that the Caribbean was biologically productive due to an abundance of nutrients derived from river discharge and upwelling. The surface area covered by this plume may exceed 160,000 km2, and therefore the only way to study it synoptically is also through the use of remote sensing.

Despite these pioneering efforts, several key questions have remained unanswered, namely:

  • (1)

    What is the nature of the plumes? Are they the result of colored Amazon water or are they the result of phytoplankton blooms stimulated by eddy pumping?

  • (2)

    How deep are the plumes in the high seas and how fast do they move?

These questions are the focus of this paper. We prove that the coloration of these waters is primarily due to the discharge of large rivers and report on their depth and dynamics.

Section snippets

Satellite data

The Sea-viewing Wide Field-of-view Sensor (SeaWiFS) satellite sensor, launched in August 1997, provides complete and calibrated coverage of the global ocean reflectance every two days (Hooker et al., 1992; McClain et al., 1998; Barnes et al., 1999). We obtained the SeaWiFS Level-3 products from the Distributed Active Archive Center (DAAC) at the NASA Goddard Space Flight Center, USA. The products included the global 9-km resolution Chl-a concentration and spectral water-leaving radiance (six

Spatial and temporal patterns

Fig. 2 presents the monthly composite of CDOM absorption coefficient distribution maps for 1997–2002. Float surface locations for each month were overlaid on the images. The colored, continuous water masses that extend from the Amazon and Orinoco River mouths (red–yellow–green colors) are the plumes or fresh-water lenses associated with the rivers. Arguments and proof for this assertion are provided below.

The CDOM absorption coefficient images (from here onward referred to as CDOM images)

Summary and conclusions

We have studied the spatial and temporal evolution patterns of colored water in the western tropical North Atlantic Ocean and in the Caribbean Sea. Using nearly concurrent (eight-day window) S-PALACE floats and SeaWiFS data collected over four years (1998–2001) we found that the colored-water mass was associated with riverine discharge originating primarily from the Amazon and the Orinoco Rivers. Several findings led to this conclusion: (1) the plumes appeared to be continuous features, linking

Acknowledgements

This work was supported by NASA grant NAG5-10738 to CH and NAG5-11254 to FMK, and NSF grant OCE-831869 to ETM and RWS. We thank the US SeaWiFS Project (Code 970.2), the Distributed Active Archive Center (Code 902), and OrbImage for the production and distribution of the satellite data. The software codes to estimate Gelbstoff and chlorophyll from SeaWiFS were provided by Kendall L. Carder (USF). We thank Robert Weisberg (USF) for helpful discussions about the oceanography of the tropical North

References (41)

  • J.F. Ternon et al.

    A seasonal tropical sink for atmospheric CO2 in the Atlantic Oceanthe role of the Amazon River discharge

    Marine Chemistry

    (2000)
  • Arnone, R.A., Martinolich, P., Gould Jr., R.W, Stumpf, R., Ladner, S., 1998. Coastal optical properties using SeaWiFS....
  • R.A. Barnes et al.

    Changes in the radiometric sensitivity of SeaWiFS determined from lunar and solar-based measurements

    Applied Optics

    (1999)
  • R.R. Bidigare et al.

    Influence of the Orinoco River outflow on distributions of algal pigments in the Caribbean Sea

    Journal of Geophysical Research

    (1993)
  • A. Bricaud et al.

    Absorption by dissolved organic matter of the sea in the UV and visible domains

    Limnology and Oceanography

    (1981)
  • A. Bricaud et al.

    Variability in the chlorophyll-specific absorption coefficients of natural phytoplanktonanalysis and parameterization

    Journal of Geophysical Research

    (1995)
  • Carder, K.L., Hawes, S.K., Lee, Z., Chen, F.R., 1997. MODIS: Case 2 chlorophyll a algorithm. MODIS ATBD-19, URL...
  • K.L. Carder et al.

    Semianalytic moderate-resolution imaging spectrometer algorithms for chlorophyll a and absorption with bio-optical domains based on nitrate depletion temperatures

    Journal of Geophysical Research

    (1999)
  • K. Ding et al.

    Analysis of the influence of O2 A-band absorption on atmospheric correction of ocean-color imagery

    Applied Optics

    (1995)
  • D.M. Fratantoni et al.

    North Brazil Current ring generation and evolution observed with SeaWiFS

    Journal of Physical Oceanography

    (2002)
  • Cited by (217)

    View all citing articles on Scopus
    View full text