Assessing modern river sediment discharge to the ocean using satellite gravimetry

Abstract

Recent acceleration of sand extraction for anthropic use threatens the sustainability of this major resource. However, continental erosion and river transport, which produce sand and sediment in general, lack quantification at the global scale. Here, we develop a new geodetic method to infer the sediment discharge to ocean of the world's largest rivers. It combines the spatial distribution of modern sedimentation zones with new high-resolution (~170 km) data from the Gravity Recovery and Climate Experiment (GRACE) mission launched in 2002. We obtain sediment discharges consistent with in situ measurements for the Amazon, Ganges-Brahmaputra, Changjiang, Indus, and Magdalena rivers. This new approach enables to quantitatively monitor the contemporary erosion of continental basins drained by rivers with large sediment discharges and paves the way toward a better understanding of how natural and anthropic changes influence landscape dynamics.

Fig. 1

Rivers which sediment discharge is estimated in this work. Gan: Ganges-Brahmaputra, Irr: Irrawaddy, Amz: Amazon, Chj: Changjiang, Ind: Indus, Mag: Magdalena, God: Godavari, Mek: Mekong, Con: Congo, Hua: Huanghe, Mis: Mississippi, Ori: Orinoco. Figure 1 created using GMT software

Fig. 2

Maps of annual offshore sedimentation and collocated GRACE signal. Modeled sedimentation rates out of a the Amazon, b the Ganges-Brahmaputra, and c the Changjiang (or Yangtze) for 10-µm diameter sediment particles. The sedimentation area contour is the pink line. d–f GRACE equivalent sediment thickness (assuming that sediment replaces water) over the same areas, the sedimentation area contour is also superimposed. Mass variations over the continents are masked to improve figure readability (see Supplementary Fig. 4 for unmasked GRACE signal over the continents). Figure 2 created using GMT software

Fig. 3

Example of the grain size control of the modeled sediment distribution at the Magdalena river. a Sedimentation area decreases when grain size increases due to a faster sink relative to the horizontal advection capability of the sea currents. b Influence of the grain size on the SSD. For the Magdalena, SSD is closer to the in situ sediment discharges when grain size is around 40–50 µm (see blue and red arrows for s1 and s2 solutions, respectively) and in any case larger than 40 µm. Results at the other 12 rivers are in Supplementary Fig. 6. The hue around each GRACE solution represents the uncertainty derived from the signal (sediment) to noise (GRACE data) analysis (see Methods)

Fig. 4

GRACE satellite-derived sediment discharge (SSD) with standard deviation vs. in situ suspended sediment discharge. For clarity, we only show positive values of SSD for grain sizes (bottom right subpanel) that better compare with in situ sediment discharge for each hydrological correction (s1 and s2). Ganges-Brahmaputra (Gan) and Irrawaddy (Irr) have negative SSD for s2 and s1 solutions, respectively. Amz: Amazon, Chj: Changjiang, Ind: Indus, Mag: Magdalena, God: Godavari, Mek: Mekong, Con: Congo