CHAPTER IV. The Work of Denudation

THE WORK OF DENUDATION

Transference of material from land to sea—The denudation ratio—Weathering of rocks—The work of chemical denudation—Summary of the data—Composition of the saline matter in the oceans and of that annually carried to the oceans—The work of mechanical denudation—Suspended and bottom loads of rivers—Mississippi not a good average case—Dole and Stabler’s work in the United States—Application to the whole land area—Total material removed and rate of degradation of land—Marine erosion—Types and quantities of sediments annually produced.

The purely geological methods which have been devised to investigate our problem are of two kinds. The first attempts to apply a time-scale to the sedimentary rocks and was, historically, the earliest to be proposed; the second, due to Joly, deals with the accumulation of salt in the oceans. The one is concerned with material carried away from the land mechanically; the other with the material removed in solution. The various agents of weathering, of which rain and frost are the chief, disintegrate the surface rocks and supply the rivers with their load of detritus. The turbid condition of rivers when in flood, heavily charged with alluvial matter, is a familiar and convincing proof that the effect of erosion in conjunction with the transporting power of running water must always be to wear down the land areas. In the dynamical study of denudation and sedimentation the first essential is to know the rate at which the rivers are working. Measurements of their load of silt and dissolved salts and of their annual discharge to the sea make it possible to arrive at reliable estimates of their activity.

Incidentally, it is useful to determine the ratio which solvent denudation bears to the whole. The denudation ratio, as it may conveniently be called, is the ratio of the load of dissolved material to the total load carried both in solution and in suspension.

The amount of material removed in solution from the surface rocks is not quite the same as that which is carried to the oceans. A small proportion is abstracted from the over-ground circulation by the waters which sink below the surface, and while some of this is undoubtedly brought up again through the agency of springs, it seems possible, as Prof. Schwarz has boldly suggested, that certain constituents, such as iron and magnesium, may be permanently removed from the earth’s crust by downward migration. In the denudation ratio this possibility is left out of account as having no bearing on the study of sedimentation.

A rough estimate of the denudation ratio may be made by considering the weathering and decay of rocks in situ. Soluble constituents are withdrawn by leaching and a residue of the more stable minerals and alteration products is left behind. Weathering involves not only the abstraction of material but also the introduction of fresh material from external sources. Oxidation, hydration, and carbonatisation are the most typical reactions, and they must be allowed for in determining the proportion of the original rock lost by solution. This can be done approximately by assuming that some element—aluminium being usually chosen—has remained invariable during the course of decomposition. From the analysis of a large number of fresh rocks and of their altered equivalents, it is found that on an average 30% is dissolved, leaving a residue of 70%. The denudation ratio ought therefore to be about 0·3.

The direct determination of the work of chemical denudation requires three distinct sets of measurements: (a) the annual discharge of rivers into the oceans; (b) the analysis of their waters; (c) the measurement of their drainage areas. Mellard Reade was the first to point out the importance of these factors in the study of dynamical geology. In 1879 he collected such information as was then available and deduced from it the quantity of rock material annually removed from the whole land area, his estimate being 5280 million tons. Sir John Murray’s corresponding estimate of 1887, which was based on analyses of the waters of nineteen of the world’s principal rivers, was 4975 million tons.

Until 1909 this figure could not be improved upon, but in that year there was published by the United States Geological Survey the results of the detailed and systematic work carried out by R. B. Dole and H. Stabler. For the first time an attempt had been made to measure the discharge, drainage areas, salinity, and suspended load of all the important rivers of a large continental area. The estimates represent the averages of observations made daily for a year or longer, and in the case of the discharge, the measurements extended over at least seven years.

Seasonal variations and the effects of floods are apt to be misleading, and a long continued series of observations is necessary if their relative importance is not to be over—or under—estimated. The amount of material in solution varies but slightly from year to year, and the average of one year’s results is within 10% of the true mean value. The annual discharge varies much more than this, and still more inconstant is the load of suspended material, which, in certain years, may differ from the average value by 50%. These figures indicate how difficult it is to introduce exact measurements into geology with any hope of finality.

A summary of all the best data now available has recently been given by Dr. F. W. Clarke, and covers about 28 million square miles of the drainage areas of the earth. The details are given in the following table:

-----------+---------------+------------------------ | | SOLVENT DENUDATION. | DRAINAGE AREA | TONS ANNUALLY REMOVED. CONTINENT. | IN SQ. MILES. +---------+-------------- | | PER SQ. | FROM WHOLE | | MILE. | AREA. -----------+---------------+---------+-------------- N. America | 6,000,000 | 70·5 | 423,000,000 S. America | 4,000,000 | 45·5 | 182,000,000 Europe | 3,000,000 | 90·0 | 270,000,000 Asia | 7,000,000 | 75·0 | 525,000,000 Africa | 8,000,000 | 40·0 | 320,000,000 | ---------- | ---- | ------------- | 28,000,000 | 61·0 | 1,708,000,000 -----------+---------------+---------+--------------

The total land area of the globe is estimated by Murray at 55·7 million square miles, but of this 11·5 million square miles are areas of internal drainage, such as the Great Basin of the United States and the Asiatic depressions, which contribute nothing to the ocean. The circumpolar regions, representing 4·5 million square miles, must also be left out of account, and the remaining 39·7, or say, 40 million square miles, is that from which the oceans are fed. If the figures given above be accepted as typical, then the annual addition of material to the oceans by solution amounts to 2440 million tons. Some of this, however, is derived from the atmosphere—chiefly as carbon-dioxide. Applying the necessary correction of nearly 10%, there remains 2220 million tons as the amount actually derived from the rocks. Murray’s estimate, it will be noticed, is almost exactly twice that of Clarke.

The composition of the saline matter carried to the oceans may be found by suitably weighing each analysis of river water according to the discharge of the latter. The general mean of all such results is given in the table opposite, together with the total amount of each substance in the ocean. The small traces of elements other than those listed are quite insignificant.

---------------------+----------------------+-------------------- | ANNUAL ADDITION | TOTAL SALINE MATTER CONSTITUENT. | OF SALINE MATTER | OF THE OCEANS IN | IN MILLIONS OF TONS .| MILLIONS OF TONS. ---------------------+----------------------+-------------------- SiO₂ | 284 | Fe₂O₃Al₂O₃ | 67 | Mg | 83 | 1,535,000,000 Ca | 497 | 490,000,000 Na | 156 | 12,616,000,000 K | 37 | 454,000,000 Cl | 138 | 22,800,000,000 Br | | 78,000,000 CO₃ | 857 | 80,300,000 NO₃ | 22 | SO | 299 | 3,172,000,000 ---------------------+----------------------+-------------------- Total | 2440 | 41,230,000,000 ---------------------+----------------------+-------------------- Annual discharge of river water into ocean = 24·3 × 10¹² tons Volume of ocean water = 307,496,000 cubic miles Density of ocean water = 1·026 (mean) Mass of ocean water = 1,178,270 × 10¹² tons -----------------------------------------------------------------

It is obvious from a comparison of these two columns that the annual increment of dissolved matter is not permanently retained by the oceans. The greater proportion is precipitated by chemical and organic agencies, and either becomes incorporated with detrital material or goes to form individual sediments. The chief substances produced are calcium and magnesium carbonate, gypsum, limonite, and silica. Rather more than one-third of the carbonates appear to associate themselves intimately with sands and muds. The greater part of the remaining two-thirds is deposited as limestone on the continental shelves (after being used by various organisms in shell-making) in waters which are comparatively free from terrigenous sediment. The abstraction of gypsum from the ocean takes place at irregular intervals under suitable conditions of concentration. Of the limonite and silica, the chief precipitation takes place on the continental shelves where they associate themselves with the detrital sediments. The history of potassium is rather obscure, but on the contrary, that of sodium appears to be the simplest of all. That it is stored up in the oceans is an assumption which is granted as justifiable by most geologists.

The only new factor required in order to estimate the mechanical work of denudation is the load of material carried by the rivers. Besides the silt transported in suspension, larger fragments are carried by rolling along the stream bottom. Measurements of the bottom load are lacking except in a solitary case—that of the Mississippi—in which it amounted to about 10% of the whole. It is difficult to define any precise difference between bottom load and suspended load, the former being only a limiting case of the latter. When the water is fully charged with rock debris the highest proportions are found near the bottom and sides, and at a point in mid-stream at about one-third the depth from the surface—the position of the stream lines of maximum velocity. In clear water the rolling power reaches its maximum value, for apart from fluid friction, energy is expended in no other way. Most rivers fall between these extremes. In making actual measurements, samples are taken from representative points in the river and used to give the average over the whole section. In the final estimate it seems probable that a large proportion of the bottom load is accounted for. Even if a correction ought to be applied it would be pedantic to do so except for rivers which have been under observation for several consecutive years, since the variation from the mean annual load is very great from year to year. The Nile varies by 40% and the Potomac by as much as 100%. In the United States the mean variation is about 50%. With uncertain data of this kind a correction of less that 10% may safely be disregarded.

Of all the rivers of the world, the Mississippi has been most favoured by measurements of the kinds required, and many estimates of the rate of continental degradation, of the rate of deposition of sediments and of the age of the earth have been based upon them. There is no doubt, however, that the Mississippi is working more rapidly than any other river of importance in North America, except, perhaps, the Colorado River. The high declivity in the west, the Tertiary elevation of the plains to which the streams are not yet adjusted, and the abundance of easily eroded glacial drift are all factors which promote this activity. In the case of rivers other than those of North America for which data are available, the same high rate of denudation obtains; the Rhone and the Po, for example, being amongst the most energetic workers in the world. Generalising for the whole earth from these rivers alone, would obviously give misleading results. The work of Dole and Stabler again comes to our aid, and in the following table their aggregate measurements for the whole area of the United States are tabulated. Similar evidence for four widely separated rivers is also given:

----------------+-----------+----------------------+---------- | | MILLIONS OF TONS OF | | | MATERIAL REMOVED | DRAINAGE BASINS.| AREA IN | PER YEAR. | DENUDATION | SQ. MILES.+-----------+----------+ RATIO. | |IN SOLUTION| IN | | | |SUSPENSION| ----------------+-----------+-----------+----------+---------- United States | 3,088,500 | 241·5 | 468 | 0·34 Mississippi | 1,265,000 | 122 | 304 | 0·29 Nile | 1,100,000 | 21 | 52 | 0·29 Uruguay | 150,000 | 7·5 | 15 | 0·33 Rhone | 34,800 | 8·5 | 36 | 0·19 ----------------+-----------+-----------+----------+----------

Leaving out the Mississippi because of its inclusion in the United States and weighing each result according to the area over which it holds, the mean denudation ratio is 0·31, a figure which agrees very well with our previous estimate. If now we use the denudation ratio to calculate the material removed by mechanical denudation over the whole land surface, we should not be far from the truth. It is clear that if from the 40 million square miles which drain into the oceans the quantity of material carried in solution represents 0·3 of the total material removed, then, as the former amounts to 2440 millions of tons annually, the quantity carried away mechanically must be 5700 million tons. That this figure is of the right order is favoured by another consideration. The mean elevation of North America is very nearly that of all the land areas of the earth. Moreover, according to Clarke’s figures the rate of denudation over North America is slightly higher than the average for all the lands, but more closely approaches it than does that of any other continental area. We may therefore take the rate of denudation of North America as a fair average and apply it with some confidence to all the drainage areas of the globe. Doing this, the total amount of suspended material annually discharged into the oceans is computed to be 6000 million tons.

We may sum up the work of denudation in round figures as follows:

Material annually removed in solution 2500 million tons Material annually removed in suspension 6000 ” ” ----------------- Total 8500 million tons

These figures may also be expressed in terms of the rate at which the land areas are being worn down. By solvent denudation a degradation of one foot in 30,000 years is implied, and by mechanical denudation, one foot in 12,000 years. Taking both together the average rate of denudation is found to be one foot in 8600 years. It should be clearly understood that individual areas may be lowered at rates very different from this. The maximum rate is attained in the Irawadi basin, one foot of which is removed in 400 years. The Po is also an exceptional river, and lowers its basin by one foot in 850 years. On the other hand, in the Hudson Bay district of North America the drainage only carries away one foot in 47,000 years.

No minimum figure can be given, for wherever deposition of sediment takes place on the land areas the temporary rate of denudation locally becomes negative. In making these calculations, the density of rock material is taken as 2·6; the weight of a cubic foot as 165 lbs; and the weight of a cubic mile as 10,800 million tons. Although the surface covering of loam or earth weighs only about 100 lbs per cubic foot, the denser and more closely packed underlying rock need alone be considered, for it is by its decay and expansion that the superficial blanket above is produced.

Our final problem is to determine the nature and quantity of the sediments which are ultimately formed on the continental shelves. This can only be done roughly, but the results will suffice to serve our purpose. First of all, two serious difficulties must be met before the way is open to take this step. So far, marine denudation has been left out of account. It is not yet possible to make a wholly satisfactory estimate of the relative magnitude of the supply of detritus captured directly by the sea. The unknown factor is the average encroachment of the sea upon the coasts. For the British Isles, Croll suggested an average of three feet per century, and the figure assumed by Sir A. Geikie about the same time was ten feet per century. A much later estimate by Prof. Watts places the average retreat of the English coast at a hundred feet per century. Along parts of our East coast marine erosion is still more rapid than this, the conditions being exceptionally favourable. On the other hand, Geikie considers that all the force of the Atlantic beating upon the N.W. coast of Scotland may not wear it away at more than one foot per century. What the average between these extremes may be can only be guessed at. If for convenience we accept Geikie’s figure as affording a likely average for all the coast lines of the earth (125,000 miles), and if the average height of the cliffs be taken as 150 feet, then the mass of material annually removed will be about 700 million tons.

The other difficulty is concerned with the annual amount of material which remains in the oceans in solution, and also of that which is deposited on the ocean floor outside the limits of the continental shelves. For the former, a knowledge of the age of the oceans is necessary. Considering all the evidence, the amount retained at the present day seems to be about 200 million tons, but this is certainly too high as a figure representing the average increase throughout the history of the oceans.

For the deep-sea deposits little more than a guess is possible, although we can now approximate to the right order of magnitude by considering the circulation of radium. The radium in the material removed from the lands is redistributed between the sediments on the continental shelves, the deep-sea deposits and the water of the oceans. Applying our present knowledge of the distribution of radium (see p. 131) the annual mass of the deep-sea deposits is found to be about ¹/₃₀ of the whole, i.e. about 300 million tons.

The difficulties can scarcely be avoided by balancing them against each other. There still remain 200 million tons (700 - 300 - 200) to be added to the 8500 million tons already found as the total for sub-aerial denudation. This gives us 8700, or as an extra safeguard, say 9000 million, tons as the mass of sediment annually deposited on the continental shelves. It is unfortunate that to arrive at this figure an element of doubt should be introduced by associating the results of careful experimental work with the vague conclusions just arrived at. The bugbear of the whole investigation is marine erosion; but if it is remembered that the figures given in that connection are meant to be suggestive rather than final, no erroneous impression need be carried away.

If the sediments ultimately formed are shales (20% quartz), sandstones (75% quartz) and limestones (75% calcium carbonate), their proportions will be as follows:

Shales 70% or 6300 million tons. Sandstones 16% “ 1440 ” ” Limestones 14% ” 1260 ” ” -------------------------- Total 100% = 9000 million tons.

In the two following chapters the application of denudational statistics to the measurement of geological time will be considered.