The Wearing Away of Continents

Abstract

Denudation is a substantial part of a cycle in which flowing water associates the terrestrial sector of the global hydrological cycle with continental wearing down, of which it is a chief agent. Tectonic uplift stimulates fluvial erosion, on contact with the water cycled by solar energy and clustered in surface channels by catchment processes. Progressive sediment transfers occur between upper and lower catchments, and subsequently between lower catchments and the marine environment. Big river flood plains store sediment in larger systems for longer periods, where reworking continues mechanical and chemical sorting before reaching the onward transfer of mature sediments to the coastal zone. This is a continuous process where coarse, raw fluvial sediments are eventually swept as molasse into trenches and back-arc basins, close to orogens. About 19.1 Gt of sediment and 3.8 Gt of dissolved phases are annually transferred to coastal oceans by river discharge. This huge amount of material is, directly or indirectly, the result of the action of weathering, which is a significant link in the cycling of carbon and, hence, a participant in controlling the Earth’s climate. A significant portion of this material is, however, an indirect product of weathering because it is rock debris that has been recycled, having passed two or more times through the Earth’s exogenous cycle. Also, anthropogenic activities are responsible for opposing actions, which increase the denudation rate through soil erosion, on one hand, and sequester sediments in human-made reservoirs, on the other.

6.1 Introduction

The synergistic action of rock weathering and erosion results in continental denudation; simply stated, continents are worn down by the disintegration of rocks, followed by the transport of the resulting debris and dissolved phases to coastal seas and oceans. Rivers are the leading actors in the play because they constitute the major transport pathways; although they only represent about 0.0001 % of the total water mass existing on the surface of the Earth, they make up an effective conveyor belt system that transports over 19 Gt of sediment and 3.8 Gt (1 Gt = 10⁹ t) of total dissolved solids (TDS) per year to the World’s oceans, thus denuding the Earth’s crust. These mechanisms of rock weathering and subsequent erosion are more complex than they appear at first sight because they not only wear away continents but also generate a rebound of the Earth’s crust that interacts with climate constituting a mechanism that intervenes in the global carbon budget, affecting the concentration of atmospheric CO₂ (e.g., Ruddiman 1997).

6.2 The Exported Products of Weathering

The series of processes that comminute and dissolve crustal minerals and rocks, as seen in previous chapters, generate products that consist of dissolved and solid phases.

The naturally occurring dissolved fraction is composed of major (e.g., sodium, calcium, silicon, magnesium, alkalinity, sulfate, etc.), minor (e.g., iron, aluminum, manganese, etc.), and trace elements (e.g., the REE, transition metals). The concentrations of major elements are determined in the range of mg L⁻¹ or mol L⁻¹; minor components are within the range of several parts per million (ppm) parts, and traces are determined in the <1 ppm (in the parts per billion parts or ppb range) or even less (e.g., ng L⁻¹). Some are soluble large cations, known as “large ion litophile” or LIL elements, such as strontium and cesium. Others are transition metals (e.g., scandium, titanium, vanadium, chromium) and high-field strength (HFS) elements, like niobium and thallium.

The solid fraction includes minerals (e.g., quartz, feldspar, micas, and heavy minerals); rock fragments; sometimes partially leached aluminosilicates (i.e., poorly ordered crystals); iron, manganese, and aluminum oxides and hydroxides; and crystalline phyllosilicates (i.e., the clay fraction).

It is not within the objectives of this brief monograph to enter the mineralogical complexities of the solid residue of weathering. The brief analysis that follows is limited to the most conspicuous characteristics of the material, whether solid or dissolved, that abandons the continents via riverine pathways.

6.2.1 Particulate Phases

The fractured material on the Earth’s surface that results from different mechanisms (i.e., mechanical weathering is significant but tectonics is also responsible for the fracture of rocks), which is amenable to additional weathering processes, varies in size from boulders (i.e., up to ~4 m in diameter) down to cobbles (256–128 mm), pebbles (64–8 mm), sand (2 to ~0.1 mm), silt (88 to ~4 μm), clay (~4–0.2 μm), and colloids (<~0.2 μm). Rock fragments and coarse rock debris are typical of high-energy systems (e.g., mountainous rivers) (Wohl 2010), whereas finer grain-size material (i.e., sand, silt, clay, and colloids) dominate in alluvial, mature rivers, with channels and floodplains that are self-formed in unconsolidated or weakly consolidated sediments. Rock fragments are generally composed of several minerals but sand and the finer grain-size fractions are usually made up by one dominant mineral phase.

Coarse geologic materials are mainly transported as bed load by high-energy rivers (also by mass movements like landslides and debris flows) and seldom reach the continental platform, although rivers draining active continental margins (e.g., along the Pacific coast of South America) are thought to be a significant conduit for the transfer of relatively coarse continental material to the sea (Milliman and Syvitski 1992). At any rate, relatively coarse grain-size sediments (i.e., sands) remain mostly along the continental–oceanic interface. Potter (1986) made an interesting study of beach sand composition in South America on the basis of the proportions of quartz, feldspar, and rock fragments (Q:F:Rf), concluding that tectonics are the dominant control on the mineralogy of beaches, showing a clear distinction between active and passive margins although, sometimes, active margin composition overprints a passive margin association, as it occurs along Argentina’s Atlantic coast. The significant role of climate becomes evident in the so-called Brazil’s Q:F:Rf association (91:4:5), which contrasts with the ratio found for the high-energy rivers draining South America’s Pacific margin (24:16:60), showing the significance of climate in the former and the dominant role of mechanical weathering in the latter.

Finer grain-size sedimentary material (i.e., fine silt, clay, and colloids) is mostly transported as total suspended solids (TSS) by large rivers to estuaries, continental platforms and, eventually, to the deeper parts of world oceans after one or more cycles of deposition and transport (e.g., by ocean currents).For several decades, the mineralogy of deep-sea sediments has been associated with continental sources (e.g., Biscaye 1965; Griffin et al. 1968), mostly supplied by rivers and, to a smaller extent, as wind-transported dust and ice-rafted debris.

Over 15 years ago, Canfield (1997) produced an interesting study on the major elements geochemistry of river particulates from continental USA. Canfield (1997) combined a model that predicted river solutes as a function of precipitation, temperature, and a restricted number of weathering parameters. With another predictive equation, he found out that the composition of river particulates depends both on climate parameters (runoff and temperature, as they play an important role in defining dissolved river geochemistry) as well as on nonclimate factors, such as elevation, relief, tectonics, and drainage basin area. To the best of our knowledge, this approach has not been tested elsewhere.

The scientific research which has been carried out for many years, seeking to assess the global geochemical nature of mineral matter exported from the continents to the seas shows that, as river drainage basins grow larger, the mean composition of their suspended load approaches the composition of the upper continental crust (UCC). Extended UCC-normalized multielemental graphs of TSS from major world rivers (Fig. 6.1) show negative departures for soluble elements (e.g., Na, Rb, Sr, Ca, Mg, and K) and enrichments for insoluble (e.g., Ti, Fe, Zr) or adsorbed (e.g., Co, Cu, Pb, Cs, and REE) elements (e.g., Depetris et al. 2003). It is interesting to add that rivers appear to preserve a geochemical source signal in fine-grained TSS just a Potter (1986) found in South America’s beach sands. The origin of sediments can be traced not only with isotopes (e.g., McLennan et al. 1993) but also with REE and other trace elements. The latter, for example, was the approach followed by Pasquini et al. (2005) when the geochemical study of the Chubut River was undertaken, finding that an active (i.e., volcanic arc) margin geochemical signature was persistent along Patagonia’s passive Atlantic coast.

Fig. 6.1

Example of an extended UCC-normalized multi-elemental graph of TSS from major world and Patagonian rivers showing outstanding negative departures which are associated with elemental solubility (e.g., Ca, Na, K, Sr, and Rb), and positive deviations which are likely related with anthropogenic enrichment (e.g., Cu, Pb, and Sb) or insolubility (e.g., Fe, Ti, and Mn). Elements are ordered to obtain a monotonic decrease of UCC abundances when normalized to primitive mantle concentrations

6.2.2 Dissolved Phases

The dissolved phases (also known as total dissolved solids or TDS) transported from the continents to the sea via rivers, groundwater, and wind-driven aerosols recognize three possible natural sources: weathering, volcanic manifestations, and ocean water recycled through atmospheric pathways. It is widely known that anthropogenic activities constitute an additional source, which may introduce human-made products in the environment or increase the flux of naturally occurring elements. Each source carries a chemical signature which, in the case of dissolved matter, is more difficult to establish than in solid phases, even in spite of the fact that the finer grain-sized material (e.g., clays and colloids) have suffered extreme modifications. The truly dissolved fraction begins with particles <1 nm (Gaillardet et al. 2005).

Meybeck (2005) has revisited the subject matter, updating many earlier studies. His main conclusions may be summarized as follows: (a) river water types are multiple; about a dozen water types can be identified according to the exposed lithology, water balance, and atmospheric input (Fig. 6.2); (b) Gibbs’ (1970) global scheme (i.e., rain, weathering, and evaporation/crystallization dominance) still holds for about 80 % of river waters but does not account for the remaining 20 %, which has more complex driving variables; (c) few areas of the world still exhibit relatively pristine river geochemistry in as much as there is a global scale increase of riverine Na⁺, K⁺, Cl⁻, and SO₄ ²⁻. On a global basis, HCO₃ ⁻ concentration appears as stable (Meybeck 2005) although some large rivers, like the Mississippi, show for the last half century an amplification in the alkalinity export rate that has been linked to increased rainfall and to changes in the amount and type of land cover (Raymond and Cole 2003).

Fig. 6.2

Piper diagram showing the chemical clusters determined by different dominant lithological compositions. The Piper classification is based on the major chemical composition of natural waters

Contrasting with major constituents, trace elements in natural waters are characterized by concentrations <1 mg L⁻¹. There are several complex factors that control the concentration of dissolved trace elements, whether natural or anthropogenic, and dealing with such subject is beyond the scope of this monograph. We can only state here that their concentrations are several orders of magnitude lower than the host rocks that brought them to the Earth’s surface (Fig. 6.3). Moreover, they exhibit an intricate dynamics which involves aqueous speciation, and adsorption onto colloids and other exchange surfaces, such as organic matter or hydrous oxides, which are the most outstanding features controlling their concentrations in the aquatic domain (Gaillardet et al. 2005).

Fig. 6.3

Example of an extended UCC-normalized multielemental graph of TDS for different dominant lithological compositions. Shaded area is determined by maximum and minimum normalized concentration for each element. UCC-normalized concentrations mostly range from 10⁻² to 10⁻⁷ and are mainly controlled by mineral solubility and adsorption processes

6.3 Material Provenance: How Much is Recycled Material?

Provenance is a term that comes from the French “provenir,” and means “to come from.” In the Earth Sciences, provenance concerns the lithologic source of a rock, usually in sedimentary rocks. It does not refer, as one would think, to the circumstances of the collection of a rock sample. In dealing with continental denudation by the action of rivers, the term is used to establish the source of the sediment load transported to coastal seas.

Most mountainous regions (e.g., the Andes, the Himalayas) have a high fraction of outcropping sedimentary rocks, which raises the potential for sedimentary-sourced particles to dominate the TSS of major rivers. Fine-grained sedimentary rocks are frequently the product of the turnover of preexistent clastic sediments because their particles mostly consist of stable weathering products, and these particles can be recycled through several episodes of burial, uplift, and erosion. The recycled condition of the TSS in major rivers has been recognized for many years but it is difficult to quantify. Veizer and Jansen (1979) estimated that, on a global basis, clastic sedimentary rocks are 65 % recycled. Notwithstanding, sediment recycling is a complex subject and the reconstruction of the burial and exhumation cycles, particularly in areas with elaborate tectonics is a problem that requires detailed analyses. As an example, Garzanti et al. (2013) have explored the complexities of sediment recycling at convergent plate margins (Indo-Burman ranges and Andaman-Nicobar ridge).

In most major rivers, there is a surplus of TSS over what would be expected from the dissolved load if only crystalline rocks are being weathered and eroded. The amount of this excess is proportional to the extent of sedimentary rock cover in the source area. This approach provides an estimate of the fraction of recycled material a river is carrying. Gaillardet et al. (1997) used this line of thought to complete a study on the chemical and physical denudation in the Amazon River basin. Later, Gaillardet et al. (1999) attempted to estimate the proportion of recycling in different drainage basins by comparing the measured TSS to that expected from the composition of the dissolved load. The results vary broadly for different rivers with values mostly falling in the range of 45–75 % recycled material. Using the described methodology, Gaillardet et al. (1999), for instance, anticipated significant recycling for the Paraná River drainage basin TSS load, estimating that about 14 % of the eroded material is supplied by pristine continental sources.

Chemical indices, as seen in Chap. 5, can assist in establishing the likelihood that a river-transported sediment load has a significant proportion of recycled material. If, for example, CIA is high and the ICV is low, then the cause could be either extensive weathering in the source area or significant recycling of previously weathered material. Conversely, if a particular TSS exhibits low CIA and high ICV, it is very likely that the parental source is composed of mostly poorly weathered crystalline rocks.

For over 20 years, geochemical and isotopic approaches to constrain sediment provenance have supplied tools that proved to be the most valuable to attain such goal. An outstanding example is the work of McLennan et al. (1993) that used trace elements (notably REE) and a myriad of isotopes (mainly from the U/Pb system) to collect fingerprints for several terrain types, recycled sedimentary rocks among them. ⁸⁷Sr/⁸⁶Sr and ¹⁴³Nd/¹⁴⁴Nd, often used with the epsilon (ε) notation, have also proved powerful methodologies to unveil sediment sources, not only in river systems, but also in wind-transported dust.

For example, we can cite the work of Depetris and Pasquini (2007), who used a Zr/Sc versus Th/Sc bivariate plot to show the possibility of significant recycling in the TSS load of upper Paraná River Andean tributaries (Fig. 6.4a). Another graph that may be useful to exhibit the likelihood of weathering and sedimentary recycling is the one that used Th versus Th/U (Fig. 6.4b). With more powerful techniques, Campbell et al. (2005) used He–Pb double dating in river-transported zircons to shed light on the recycling and provenance aspects of the Ganges and Indus rivers. They found, for example, that at least 60 % of the Indus and 70 % of the Ganges detrital zircons have been recycled from earlier sediments. Furthermore, the results implied that ~75 % of the eroded material from the Himalayas is derived from areas of anomalously high erosion, with the short-term exhumation rate exceeding the long-term average.

Fig. 6.4

a Th/Sc ratio and Zr/Sc ratio increase from mafic to felsic source areas. The anomalous Zr is related to recycling of older sediments. Data from the indicated sources; some Paraná tributaries (Bermejo and Paraguay) show the likelihood of recycling. b Plot of Th/U versus Th for TSS of Paraná River and tributaries. During weathering, there is an increase trend of Th/U above UCC values (3.5–4.0). A similar example was shown by McLennan et al. (1993)

6.4 Weathering, CO₂ Consumption, and Climate Change

As seen in Chap. 4, the action of weathering is an effective mechanism to sequester CO₂. In a worthwhile revisit of this aspect we can recur to the following example; in the presence of carbonic acid, the Mg-olivine forsterite decomposes into:

$$ {\text{Mg}}_{2} {\text{SiO}}_{{4({\text{s}})}} + 4{\text{CO}}_{{2({\text{g}})}} + 4{\text{H}}_{2} {\text{O}} \to 2{\text{Mg}}_{{({\text{aq}})}}^{2 + } + 4{\text{HCO}}_{{3({\text{aq}})}}^{ - } + {\text{H}}_{4} {\text{SiO}}_{{4({\text{aq}})}} $$

(6.1)

Four moles of CO₂ are taken from the bacteria-supplied soil air, of which two are immobilized when Mg²⁺ is precipitated as magnesite:

$$ 2{\text{Mg}}_{{({\text{aq}})}}^{2 + } + 4{\text{HCO}}_{{3({\text{ac}})}}^{ - } \to 2{\text{MgCO}}_{{3({\text{s}})}} + 2{\text{CO}}_{{2({\text{g}})}} + 2{\text{H}}_{2} {\text{O}} $$

(6.2)

It is clear then that for every mole of weathered forsterite, two moles of CO₂ from the gaseous pool become fixed in the form of carbonate and the other two CO₂ mol return eventually to the atmosphere. The implications of this simple geochemical mechanism have been treated extensively by many authors (e.g., Edmond and Huh 1997), underlining the role played by the weathering-connected carbon dioxide in the control of the Earth’s climate.

The described mechanism is only a link in the chain of climate control. For example, increased erosion of high-elevation regions causes a rebound (isostasy) of the underlying rock layers in response to overburden removal. However, the resulting uplift compensates only partially the height-loss due to erosion, thus leaving the original Earth’s surface lower than it had been. This system driven by world tectonics has been also linked to climate change (e.g., Ruddiman 1997) and supplies an intricate vision of the scenario that surrounds continental denudation.

It is clear then that weathering consumes CO₂ (e.g., Boeglin and Probst 1998) and, as long as there is an association with a viable mechanism, it may insure the sequestration of carbon.

If, in contrast, the evasion of CO₂ from inland waters is not significantly hindered, then respired CO₂ originated largely from terrestrial ecosystems (i.e., introduced into the hydrosphere as dissolved soil CO₂), the oxidation of organic carbon, the acidification of buffered waters, etc., determines a significant efflux of CO₂. Modeling mineral weathering with the PHREEQC computer code in a high mountainous river (Córdoba, Argentina) has shown the significance of CO₂ in the overall weathering mechanism: in the stream’s uppermost reaches, CO₂ accounts for nearly 57 % of the total phases intervening in the process (Lecomte et al. 2005). Adding significantly to the whole picture, recent research has shown that a large set of streams and rivers in the USA are supersaturated with CO₂ when compared with the atmosphere, emitting 97 ± 32 Tg (1 Tg = 10¹² g) of carbon each year (Butman and Raymond 2011).

The sequestration of atmospheric carbon has played a main role in the theories that propose such mechanism as the trigger that determined periods with long-term CO₂ changes (i.e., with low temperatures). Such is the case of the hypothesis identified with an acronym based on the initials of the authors (BLAG) (Berner et al. 1983) and, years later, the so-called Raymo hypothesis (Raymo and Ruddiman 1992). Both hypotheses propose dissimilar mechanisms to control low-temperature periods in the geological past although they both assign weathering and carbon sequestration a relevant role.

6.5 Continental Runoff

Runoff is understood as the water flow that occurs when the surface (i.e., soil, sediment, or rock) is permeated to its full capacity and the excess water from rainfall, melt water, or other sources flows over the land. Runoff is relatively short-lived and soon ends reaching the channel, which is the passageway of a river. Responsive to variations in water discharge and load (i.e., solid and dissolved), the channel constantly adjusts its shape and course, thus constituting a dynamic element of the landscape. Runoff is thus transformed into stream flow.

A river’s channel is an efficient means for running water. Hence, the first requirement that affects runoff and stream flow within a basin would be the supply of water, making climate an important item in any river scenario. It eventually determines not only the water supply through atmospheric precipitation, but also the degree to which that precipitation is returned to the atmosphere as water vapor before it can contribute to stream flow. Climate is considered to include not only the long-term moisture supply, but also the daily weather patterns that establish the timing and total quantity of water supplied to a drainage basin. Controlling weather variability there are intervening factors such as storm types, typical duration and intensity of rainfall, frequency, distribution of precipitation over the basin, kind of precipitation (rainfall or snowfall), and reliability of precipitation over a period of years. Periodic floods are frequently influenced by physical conditions within the drainage basin, but droughts are controlled entirely by climate.

The riverine network that dissects the Earth’s surface provides the key linkage between land and sea, delivering yearly about 36,000 km³ of freshwater to the coastal seas. Although impressive, this enormous volume represents only ~0.0001 % of the total water at the Earth’s surface. As much as 40 % of total riverine discharge is supplied only by a dozen big rivers, like the Amazon, Congo, Orinoco, Changjiang, etc.

Climate, volcanism, modifications in the relative base level but, above all, plate tectonics are jointly responsible for the location, size, shape, and orientation of big river systems. Most of today’s large river systems in Eurasia and the Americas date from the Miocene; some are much older and originated in the Mesozoic and some as far back as the Carboniferous. A few, such as China’s Yellow and Yangtze are thought to have a Pliocene–Pleistocene age, and the present Nile was formed in the Pleistocene. The Paraná and Uruguay rivers, for example, are thought to have originated during the Late Cretaceous but the Paraguay River, a significant tributary to the former, is a younger feature, with likely Mid-Tertiary age (Potter and Hamblin 2006).

Having examined briefly the question of the geological age of the current riverine network, we can now return to the aspect of continental runoff. River discharge is—as stated above—a direct consequence of climate variability and the size of the drainage basin. Data collected from 1,100 world rivers have shown that, as expected, large drainage basins usually have higher discharges (Milliman and Farnsworth 2011). Similar drainage areas and comparable climate often exhibit similar discharge. For example, in Argentina’s arid Patagonia, the Deseado (14,450 km²) and Coyle (14,600 km²) rivers have equivalent mean discharges (~5 m³ s⁻¹) (Depetris et al. 2005). There are, however, other examples that show the significance of the reverse: in comparable drainage basin areas but contrasting climate, discharge can vary by two to three orders of magnitude (Milliman and Farnsworth 2011).

Let us use South America’s Paraná River as an example to illustrate some peculiarities of big rivers’ hydrology. The Paraná River, in terms of drainage area, is the fifth largest basin in the world (~2.6 × 10⁶ km²). Near its mouth, the Paraná joins the Uruguay River to make up the Río de la Plata drainage basin. Added together the two drainage systems reach a total area of 3.17 × 10⁶ km² and a mean discharge of ~21,500 m³ s⁻¹ (Pasquini and Depetris 2007 and references therein).

With ~180 mm y⁻¹, Paraná’s runoff is relatively low for global standards (the Amazon’s runoff is ~1,000 mm y⁻¹), and its current mean discharge is ~500 km³ y⁻¹ (~17,000 m³ s⁻¹), a water flow that places the Paraná within the World’s twelve topmost discharges. Currently, most of Paraná total annual discharge (~63 %) is supplied by the upper Paraná River, which has its uppermost catchments near the Brazilian city of Sao Paulo. Figure 6.5a shows the over 100 years-long (1904–2011) mean monthly discharge time series at the gauging station of Corrientes, ~1,200 km upstream the river mouth. There are three features which become promptly apparent: (a) the series shows a marked seasonality (i.e., high waters occur during the austral summer and low waters during winter); (b) there is a pronounced variability in mean monthly discharges; (c) interspersed in the series there are outstanding flood events. Since the series exhibits a regular seasonal fluctuation then, for the purpose of analysis (i.e., to estimate an underlying trend) it is necessary to remove the seasonality to produce a deseasonalized dataset. In the Paraná, the series with suppressed seasonality shows marked positive departures particularly noticeable in the period after 1970 and what appears to be more frequent negative departures prior to that year (Fig. 6.5b).

Fig. 6.5

a Paraná River monthly mean discharge series at Corrientes (~1,200 km upstream from the mouth) for the period 1904–2012. The series is characterized by seasonality and unusual discharge events. b The same as a but deseasonalized. Positive departures are frequently associated with El Ñiño events whereas negative are often linked with La Niña episodes

The 1904–2012 deseasonalized series can be used to scrutinize for recurrent signals by means of a harmonic analysis technique. In this case, as in earlier instances (e.g., Pasquini and Depetris 2007; Depetris and Pasquini 2008), we used the continuous wavelet transform (CWT) approach, which has the advantage of localizing the time-scale of a signal, revealing trends, breakdown points, and discontinuities (e.g., Nakkem 1999). Figure 6.6a shows the wavelet spectra for the 2 to 6-year period, computed for the deseasonalized data of Fig. 6.5b. The power spectra show a clear signal supplied by high transform coefficients (power) for quasi-decadal periods and, also, in the El Niño-Southern Oscillation (ENSO).

Fig. 6.6

a Real part of the continuous Morlet wavelet spectrum of the Paraná River deseasonalized monthly mean discharge at Corrientes (1904–2012); the framed areas show the occurrence of decadal and interannual periodicities in ENSO frequencies. b Wavelet power frequency-range for the 2–6 years frequency band that shows outstanding increased integrated power for the 1982–1983 and 1997–1998 El Niño events. Very strong El Niño episodes are identified with red arrows and strong La Niña events with blue arrows

Figure 6.6b shows the power spectra for the 2 to 6-year frequency range, computed for the deseasonalized Paraná River discharge series (Fig. 6.5b). Clearly, this range includes the period of recurring ENSO events, exhibiting the impact of the strong 1983 and 1997 occurrences.

It is not within the objectives of this monograph to enter the detailed analysis of global runoff (or river discharge) going beyond the illustrating examples used above. Milliman and Farnsworth (2011) have produced an updated, expanded, and valuable state-of-the-art. As a sum up, the global distribution of precipitation determines that South America’s large rivers and those draining Southeast Asia and Oceania account for almost 70 % of the freshwater discharge to the continental coasts (Fig. 6.7; Milliman and Farnsworth 2011).

Fig. 6.7

Latitudinal allocation of global precipitation (bold line) and in different continents. Note the substantially lower Euro-African precipitation in the ~20° N–20° S latitudinal band (Milliman and Farnsworth 2011). Reproduced with permission, Cambridge University Press

Rivers must be examined within the context of their history. This is the reason why relatively long series of discharge data are useful to assess discernible changes in their mean trend; the longer the series, the more confidence in the statistically significant results. Let us again use information recorded in the Paraná River to examine if its deseasonalized series shows a significant trend. Figure 6.8 shows the trend analysis performed by means of the Mann–Kendall methodology (also known as Kendall τ), which clearly shows a statistically significant positive slope, thus indicating that mean annual discharges have been increasing consistently, particularly during the last 50 years (Pasquini and Depetris 2007).

Fig. 6.8

Mann–Kendall trend test for data in Fig. 6.5 (1904–2012). Paraná’s mean annual discharge has increased ~18 % during the last 100 years at a mean rate of ~30 m³ y⁻¹

The seasonal Kendall test was employed to examine monthly trends. Table 6.1 shows the levels of significance that identify increasing discharge trends for the July–December period (i.e., austral winter and spring). It can be stated, then, that the Paraná’s overall increasing discharge trend is basically accounted for by the increasing precipitation in the drainage basin that occurs in the austral winter and spring. The rest of the year, the river discharge does not show an apparent modification of its regime.

Table 6.1 Seasonal Kendall test results of the Paraná River monthly mean discharge time series (1904–2012) at Corrientes

After considering unusual climatic events that may determine very high or very low runoff, a query immediately arises: what is the impact of such extraordinary events on the wearing down of continents? There is no simple answer to this question. In some fluvial systems, episodic events can occur over very short time intervals, frequently hours to weeks. The direct example is the seasonal flood during which extraordinary amounts of water and sediment are discharged by rivers. This situation is especially true in small- medium-sized mountainous rivers and streams. In larger fluvial systems, the flooding season can last months and the linkage process-response may be more intricate. Hurricane and typhoons, due to their catastrophic characteristics, can play greater roles in the discharge of water and sediment.

Concluding, we can state here that rivers can be conceived as integrators of the opposing effect of precipitation and evapotranspiration, and that they are dynamic systems whose runoff is far from stable, often subjected to events that profoundly impact on their regimes.

6.6 Continental Denudation

Denudation comes from “denudare,” the Latin word that means “to strip of all covering, to lay bare.” In the Earth Sciences, it is the term used to identify the long-term effect of all the exogenous processes added together, which cause the wearing away of the Earth’s surface, leading to a decrease in elevation and relief of landforms and landscapes. Endogenous processes such as plate tectonics, and the related volcanic activity and earthquakes, uplift and exhume continental crust exposing it to processes linked to denudation, such as weathering, erosion, and mass wasting. As seen in previous chapters, the action of mineral and rock weathering results in a solid and dissolved residue and we will consider now the continent-ocean transfer separately.

The most updated database available to compute solid continental denudation via riverine transport to the Earth’s oceans, supplies a mass transport rate of ~19.1 Gt y⁻¹ (or 19.1 10⁹ t y⁻¹), which can be translated to a global mean sediment yield of 190 t km⁻² y⁻¹ (or ~0.1 mm y⁻¹) (Milliman and Farnsworth 2011). Earlier computations arrived at 18.3 Gt y⁻¹ (Holeman 1968), 13.5 Gt y⁻¹ (Milliman and Meade 1983), and 16 Gt y⁻¹, using an empirical approach (Ludwig and Probst 1998). These calculations are based on numerous determinations of the mass of sediments transported in suspension (TSS) or as bed load along the river bottom. The TSS load includes the wash load, which is the fine grain-size fraction that is mostly transported in suspension, and the bed material load, which is the sediment incorporated into the TSS during higher discharge. As river flow decreases, the sediment that is normally transported as bed material load may be incorporated to the river bed load. The opposite is also true: increased flow may incorporate a fraction of the bed load into bed material load.

Measuring the transport rate of the sediment bed load in a river, particularly if it is large, is a difficult task with traditional invasive methodology (i.e., sampling devices that are placed on the river bottom). Modern techniques, such as acoustic Doppler current profiler (ADCP), have arrived to supply a much less intrusive approach. However, most workers still assume that the bed load fraction of the total sediment transported by rivers is relatively small (e.g., ~10 %), although this relative amount may be too high for large meandering rivers or too low for steep, mountainous rivers, where the bed load mode of sediment transport operates almost exclusively during extreme flooding events.

Some of the complexities that surround obtaining an accurate measure of the sediment delivered by large rivers to the coastal ocean are illustrated in Fig. 6.9. A well-defined, confined channel, as in the Upper Paraná River (Fig. 6.9a) allows determining a much more reliable relationship, in statistical terms, between discharge and sediment transport rate. In a river section as the one shown in Fig. 6.9b, the mentioned relationship tends to be blurred by the water exchange that occurs between numerous side channels, the wide flood plain with ox-bow ponds and other morphological features, on one hand, and the river’s main stem, on the other.

Fig. 6.9

a Relationship between discharge and sediment transport rate in the Upper Paraná River at Posadas (~1,500 km upstream the mouth); note in the Google Earth image the narrow flood plain, with high banks, and confined river channel. Measured suspended sediment concentrations can be correlated with river discharge to derive a sediment-rating curve. b Same as a but at Paraná (~400 km upstream the mouth); note the channel braided pattern and the wide (~30 km) flood plain, with ponds and secondary channels

Again following Milliman and Farnsworth (2011), the above-stated global sediment flux of 19.1 Gt y⁻¹ is distributed, in a descending order, as follows: Oceania ~37 %; Asia ~28 %; South America ~12 %; North/Central America ~10 %; Africa ~8 %; Europe ~4 %; and Eurasian Arctic ~1 %. These results underline the role of precipitation (i.e., climate) and relief (i.e., partly a surrogate for tectonics) in defining significant denudation and showing, as well, the linkage with weathering (Fig. 6.10).

Fig. 6.10

Annual discharge of TSS to the global coastal Ocean. Fluxes in Gt y⁻¹. The total sediment mass is ~19 Gt y⁻¹. The figure was obtained from Milliman and Farnsworth (2011). Reproduced with permission, Cambridge University Press

As seen in earlier chapters, weathering generates a solid and a dissolved residue. Rivers, therefore, transport not only particulate matter but also the TDS that result from the action of weathering. In a schematic representation of the weathering reaction of a plagioclase, like the one seen in Chap. 4 (e.g., Na-plagioclase + water + carbonic acid → kaolinite + sodium ion + alkalinity + dissolved silica), a significant percentage (e.g., 20–30 %) of the initial reactants are transferred to the solution and, therefore, exported from the system. Therefore, the appraisal of denudation implies the consideration of the dissolved solids that are removed from the continental mass and transported to world oceans.

There has been considerable debate over the supremacy of climate over lithology (or the other way around) in the control of the riverine chemical signal. However, the idea that lithology is the factor that rules the nature and quantity of the TDS efflux is currently gaining ground. Climate, on the other hand, appears to control the rate of chemical weathering and, hence, of dissolved denudation.

World wide data gathered on the total dissolved mass transport rate shows that rivers discharge annually about 3.8 Gt of dissolved phases to the ocean, the greatest contribution being delivered by Asian rivers, which supply ~37 % of the global flux (Fig. 6.11). Europe occupies the second place with ~13 %, and the Amazon runs third with ~10 % (Milliman and Farnsworth 2011).

Fig. 6.11

Riverine delivery of TDS to the global coastal ocean. Numbers are mean concentrations in mg L⁻¹. Circles represent mean concentrations for selected rivers. Estimated global continental flux is ~3.8 Gt y⁻¹. More details are provided in the book by Milliman and Farnsworth (2011). Reproduced with permission. Cambridge University Press

Before concluding the abridged treatment of continental denudation, we must consider, albeit briefly, the anthropogenic role on the denudation scenario. It is known that human activities have concurrently increased the sediment transport by global rivers through soil erosion—estimated by Syvitski et al. (2005) in 2.3 ± 0.6 Gt y⁻¹- and, at the same time, decreased the sediment transfer of particulate material to ocean coastal areas by retaining 1.4 ± 0.3 Gt y⁻¹ in reservoirs (Syvitski et al. 2005). The authors have estimated in this significant paper, that over 100 Gt of sediment and 1–3 Gt of carbon are presently sequestered in reservoirs worldwide, which were mostly built within the past 50 years.

Sediment loads and yields have been frequently translated into denudation rates (e.g., Judson and Ritter 1964). Presupposing an average specific gravity of 2.0 for the upper portion of the crust that is subjected to erosion, a sediment yield of 2,000 t km⁻² y⁻¹ is approximately equal to a denudation rate of 1 mm y⁻¹. Therefore, yields that fluctuate between 100 and 10,000 t km² y⁻¹ can be translated into 0.05 and 5 mm y⁻¹. These rates, expressed as a uniform decrease in height of the landscape, do not mean, naturally, that denudation occurs in this manner, but it is a simple way to compare rates.

Glossary

Bed load:

The term is used to describe the transport of sand, gravel, boulders, or other debris by flowing water by rolling or sliding along the bottom of a stream.

Colloid:

It is a solid or liquid substance microscopically dispersed throughout another substance (e.g., water). The dispersed-phase particles have a diameter between ~1 and ~1,000 nm (1 nm = 10⁻⁹ m). The dispersed-phase particles or droplets are affected largely by their surface chemistry.

Comminute:

From comminution; the process in which solid materials are reduced in size, by crushing, grinding, and other processes. It occurs naturally during faulting in the upper part of the Earth’s crust, and it is an important unit operation in mineral processing, ceramics, electronics, and other fields.

Continuous wavelet transform (CWT):

It provides a redundant but detailed description of a signal in terms of both time and frequency. It is used to divide a continuous-time function into wavelets. Contrasting with Fourier transform, the CWT is characterized by the ability to construct a time–frequency representation of a signal that offers very good time and frequency confinement.

Deseasonalize:

In statistics, data are deseasonalized when the regular seasonal fluctuations are removed from a time series.

El Niño-Southern Oscillation (ENSO):

It is a band of anomalously warm ocean water temperatures that occasionally develops off the western coast of South America and can cause climatic changes across the Pacific Ocean. The Southern Oscillation refers to variations in air surface pressure in the tropical western Pacific and in the temperature of the surface of the tropical eastern Pacific Ocean (warming and cooling known as El Niño and La Niña, respectively). The two variations are attached: the warm oceanic phase, El Niño, occurs with high air surface pressure in the western Pacific, while the cold phase, La Niña, accompanies low air surface pressure in the western Pacific.

Epsilon (ε) notation:

It is an alternative way of expressing isotope ratios which allows greater flexibility in the way in which isotopic data are presented; the value is a measure of the deviation of a sample or sample suite from the expected value in a uniform reservoir and may be used as a normalizing parameter for samples of different age. It is normally calculated for ¹⁴³Nd/¹⁴⁴Nd ratio and is used to represent parts in 10,000 by the following equation:$\varepsilon^{143} {\text{Nd}} = \left[ {\frac{{\left( {\frac{{143{\text{Nd}}}}{{144{\text{Nd}} }}} \right){\text{sample}}}}{{\left( {\frac{{143{\text{Nd}}}}{{144{\text{Nd}} }}} \right){\text{standard}}}} - 1} \right] \times 10,000$

Exogenous cycle:

It is a set of events or processes which are completed, returning to its beginning and then repeating itself in the same sequence. It describes the fluxes of materials, water, and gasses that occur at the intersection of the lithosphere, biosphere, hydrosphere, and atmosphere.

Harmonic analysis:

A branch of mathematics concerned with the representation of functions or signals as the superposition of basic waves, and the study of and the generalization of the notions of Fourier series and Fourier transforms, harmonic functions, trigonometric series, almost periodic functions, and others.

Isostasy:

The term is used in geology to refer to the state of gravitational equilibrium of all large portions of Earth’s lithosphere as though they were floating on the denser underlying layer, the asthenosphere, a section of the upper mantle composed of plastic rock that is about 110 km below the surface.

Molasse:

The term refers to the sandstones, shales, and conglomerates formed as terrestrial or shallow marine deposits in front of rising mountain chains. These deposits are typically the nonmarine alluvial anyd fluvial sediments of lowlands, as compared to deep-water sediments.

Phyllosilicates:

A class of silicate minerals that form parallel sheets of silicate, where a central silicon atom is surrounded by four oxygen atoms at the corners of a tetrahedron. Three of the oxygen atoms of each tetrahedron are shared with other tetrahedrons. Examples are the clay minerals, kaolinite, and illite.

Total dissolved solids (TDS):

Are a measure of the combined content of all inorganic and organic substances contained in water, as molecular, ionized, or micro-granular (colloidal sol) suspended form. The operational definition is that the solids must be small enough to survive filtration through a filter with 2 μm nominal size pores (or smaller). TDS are usually discussed only for freshwater systems.

Total suspended sediment (TSS):

The portion of the sediment that is carried by a fluid flow, such as in a river or coastal current. It is maintained in suspension by turbulence in the flowing water and consists of particles generally of the fine sand, silt, and clay size.