A sediment budget for the Amazon basin

Cosmogenic nuclides are the state-of-the-art tool for geomorphologists: they serve to quantify the rate at which mountains are ground down by erosion, for example. To do so, geomorphologists sample a handful of river sediment from a mountain stream and measure the so-called in situ-produced cosmogenic nuclide Beryllium-10 (10Be) in the mineral quartz. But what happens to this erosion rate signal when such sediment is stored for hundreds to thousands of years in the basins and floodplains surrounding mountain belts? (ref 1) Research of our group marks the first application of cosmogenic nuclides to large lowland basins. We focus on the largest river system on the planet- the Amazon basin. In this system, roughly one billion tons of sediment are carried from the Andes to the Atlantic Ocean every year, an amount that contributes significantly to the global sediment budget. By using cosmogenic nuclides, we were able to “fingerprint” the erosion signal that moves through the lowland basin- all the way from the Andes to the ocean, a distance of roughly 3,000 km (ref 2). From measurements of a few thousand atoms of cosmogenic 10Be, we have determined the erosion rates of all mountains supplying sediment to the Amazon basin, which, when multiplied with these mountains’ area provides us with so-called sediment production rates. Such sediment production rates (shown in Figure 1A) are needed when a sediment budget of a basin is to be constructed. However, the time period over which this budget is constrained depends on the methods used. Cosmogenic nuclides cover time periods of several thousand years. We can compare this with recent measurements of the sediment flux by hydrologists. These measurements span a few decades. Comparison between both methods allows us to decipher if sediment in a basin is mostly exported to the ocean by erosion or is mostly stored internally due to depositional processes. Our group found that the same amount of sediment that enters the basin from the Andes leaves the basin at its mouth (see Figure 1B). This could mean that Amazonian sediment fluxes reaching the ocean evidently have not changed over the past several thousand years. An explanation for this trend lies in the nature of the lowland basin itself: the Amazon floodplain system is so large that small (compared to the billion tons of sediment) or very young changes (such as those due to human impact) do not result in a direct response. Instead, the Amazon floodplain seems to acts as a large buffer for eroded sediments.

Another study by our group (Ref. 3, Fig. 4) highlights the complexity of the floodplain-river system, where sediments in the modern river channel are continuously exchanged with old, highly weathered floodplain material. We quantified the mixing of old floodplain versus “young” Andean sediment, carried in the main channel, by using a cosmogenic nuclide pair (26Al and 10Be). The differential decay of these two radioactive cosmogenic nuclides helped to assess an “age” of ancient Amazonian floodplain sediment. Our results show that in some areas the Amazonian floodplain sediment is surprisingly old: in the central floodplain of the Amazon River, sediment that enters the main stream channel from some tributaries has previously been deposited and preserved from exchange with the Amazon River for between 0.5 to 3 million years.


Figure 1: Sampling device for measurements of sediment concentrations in large rivers

Figure 2: Measured sediment fluxes (all values in Mt/yr) from in situ-produced cosmogenic nuclides. Red arrows indicate the Andean sediment production, which amounts to 610 Mt/yr. The sediment flux exported at the Amazon River mouth amounts to also 610 Mt/yr from cosmogenic nuclides, and the cratonic shields located north and south of the main Amazonian lowlands, contribute negligible amounts of sediment.

Figure 3: Measured sediment fluxes (all values in Mt/yr) from recent river gauging. Red arrows indicate the Andean sediment production, which amounts to 850 Mt/yr. The sediment flux exported at the Amazon River mouth amounts to also 1050 Mt/yr from gauging, and the cratonic shields located north and south of the main Amazonian lowlands, contribute negligible amounts of sediment.

Figure 4: 26Al-10Be System: (left) All samples that plot on the grey area have never been modified by burial in the floodplain and thus represent recent, young sediment from the Andes (samples in black). The ratio of 26Al/10Be represents that of the production rates of the two nuclides. (right) White samples plot on a mixing line between young and old sediment (in red) from the floodplain.

Recent and paleo-sediment fluxes in the Amazon basin using “meteoric 10Be”

The budgeting of sediment fluxes from cosmogenic nuclide concentrations is not only possible on short time scales, but for the case that suitable archives are present, the method provides a state-of-the-art tool to quantify erosion over millions of years, up to the Pliocene. Within the Clim-Amazon project we investigate the erosion history through the Quaternary and late Neogene in the Amazon basin. Since the Amazon River is one of the main sediment providers globally, it contributes significantly to the global sediment budget and thus offers the unique possibility of being able to understand the feedbacks between tectonics (via the mass of sediment being weathered) and climate (via the consumption of atmospheric CO2 from weathering). With depths of several hundreds of meters, core samples within the Amazon basin and the Amazon fan provide a unique sedimentary archive to decipherer changes in erosion rates through time. These archives are sampled and the in situ-produced 10Be and 26Al concentrations are measured from quartz minerals, which then can be transferred to sediment fluxes. From the ratio of 26Al to 10Be, also an age can be assessed from the differential decay of the two nuclides (see Project 1, Fig. 4).

A different methodological approach is the meteoric 10Be technique that focuses, in contrast to the in-situ method (10Be produced within the mineral grain), on 10Be produced in the atmosphere. This method can be used to determine basin-wide erosion rates and weathering intensities by measuring the 10Be/9Be ratio on a soil or water sample. This new tool will be explored in recent Amazonian suspended matter and filtered water. The 10Be produced in the atmosphere is scavenged by rainfall on the Earth’s surface. Once delivered to the solid and aqueous reservoirs (see Fig. 5), it is mixed with the stable counterpart 9Be which is released from the weathering zone. Be is highly reactive and tends to be adsorbed to clays and hydroxide mineral surfaces. In rivers, when reactive and dissolved Be equilibrate, a catchment-wide erosion rate can be determined. This approach has been tested for reactive Amazon sediment (Ref. 4) with very promising results: The calculated erosion rates of 0.01 mm/yr for the geological stable shields and 0.5 mm/yr for the Andes compare well to erosion rates from in-situ produced cosmogenic 10Be (see Project 1 and Fig. 6).

Figure 6: Erosion rates εmeteo (mm/yr) of the Amazon basin from meteoric 10Be, measured on reactive sediments, compared to in situ-derived erosion rates εinsitu (see Project 1).



1) Wittmann, H.; von Blanckenburg, F. (2009): Cosmogenic nuclide budgeting of floodplain sediment transfer. Geomorphology, 109, 3-4, 246-256.

2) Wittmann, H.; von Blanckenburg, F.; Maurice, L.; Guyot, J.-L.; Filizola, N.; Kubik, P. W. (2011): Sediment production and delivery in the Amazon River basin quantified by in situ–produced cosmogenic nuclides and recent river loads. Geological Society of America Bulletin, 123, 5-6, 934-950.

3) Wittmann, H.; von Blanckenburg, F.; Maurice, L; Guyot, J.-L.; Kubik, P. W. (2011): Recycling of Amazon floodplain sediment quantified by cosmogenic Al-26 and Be-10. Geology, 39, 5, 467-470.

4) Wittmann, H.; von Blanckenburg, F.; Bouchez, J.; Dannhaus, N.; Naumann, R.; Christl, M.; Gaillardet, J. (2012): The dependence of meteoric 10Be concentrations on particle size in Amazon River bed sediment and the extraction of reactive 10Be/9Be ratios. Chemical Geology, 318-319, 126-138.

5) von Blanckenburg, F.; Bouchez, J., Wittmann, H.; Earth surface erosion and weathering from the 10Be(meteoric)/9Be ratio. EPSL, 351-352, 295-305.


Profile photo of  Dr. Hella Wittmann-Oelze

Dr. Hella Wittmann-Oelze
Geochemistry of the Earth's surface

Building E, room 228
14473 Potsdam
tel. +49 331 288-2820