In this article we shall discuss the action of rivers on the landscape, and show how the characteristic sediments deposited by them can allow us to identify ancient river courses in the geological record. The subject of river deltas will be dealt with in a subsequent article.
The reader should note that there is no qualitative difference between a stream and a river; a stream is simply a small river, or, to put it another way, a river is a big stream. Rather than write "rivers or streams" over and over again, we shall write about rivers, and the reader may assume that what we have to say applies on a smaller scale to streams.
Braided and meandering rivers
Rivers, unless artificially banked, rarely flow in completely straight lines. At low gradients, two characteristic forms they can take are braided and meandering rivers.
Braided rivers, as the name suggests, consist of a number of channels which separate and rejoin around bars of sediment. They are formed when a river with a lot of sediment repeatedly deposits the sediment and erodes it; so the braids and bars (i.e. the ridges of sediment breaking the surface of the water) are not permanent features, but shift around over time.
Meandering rivers wind from side to side in large loops (meanders). As a consequence of the hydrodynamics of this situation, the current is faster and causes more erosion on the outside of the loop, while sediment will tend to be deposited on the inside of the loop, forming point bars. The result of this is that a meandering river will become more meandering over time.
If this tendency goes far enough, the meander approaches a loop doubling back on itself. Eventually the meander may touch itself and the river will suddenly find itself with a new, straighter, shorter path, leaving the meander isolated as an oxbow lake.
The shifting of meandering and braided rivers across the landscape as a result of their own deposition of sediment produces a flattened, sediment-rich landscape known as a flood plain.
The photograph to the right shows the Rio Negro as seen from space: as you can see, the landscape is scarred with former meanders where the river used to flow before it shifted its bed. Note also the oxbow lake: its peculiar orange color is probably caused by salt-loving bacteria in the water, and should not be taken as being typical of oxbow lakes.
The sedimentary structures formed by a river at a particular point will depend on its velocity, its depth, and the sediment type. Geologists can discover the relationship between these factors both by observing actual rivers, and by laboratory experiments using flumes, allowing them to control variables such as the velocity of flow and the size of the clasts.
For example, consider the effect that a river's velocity has on a bed consisting of average-sized sand grains. At low velocities, the creep of sand along the bed will, if anything, tend to smooth out the bed.
At higher velocities, sand ripples begin to form: small ridges of sand with the ridge at right-angles to the current. These ripples have a characteristic profile with a shallow slope on the upstream side and a steeper slope on the downstream side. Saltation bounces particles of sand up the shallow upstream side and over the peak of the ripple, eroding the upstream side and depositing sand on the downstream side. This has the effect that the ripples march downstream; it also produces cross-bedding.
At a higher velocity still, dunes (essentially, big ripples) will form; as with ripples, they have a shallow slope on the stoss side and a steeper slope on the lee side. Dunes formed in this way often have ripples on their shallow stoss side; these are known, logically enough, as rippled dunes. As with ripples, transport by saltation moves the dunes and ripples downstream (with the ripples moving rather faster than the dunes) and produces cross-bedding.
At greater velocities still, especially when the sand is fine, the increased current will flatten out the ripples, resulting in a flat, layered surface known as an upper plane bed.
At still higher velocities, antidunes form. These have a rounded undulating cross-section. While in dunes sand is eroded from the upstream face of the dune and deposited on its downstream face, in the case of antidunes, sand is eroded from the downstream side of the antidune and deposited on the upstream side of the next antidune downstream. This has the effect that although the sand is moving downstream, the antidunes, being eroded on their downstream sides and built up on their upstream sides, move upstream: this is why they are called antidunes. These may show some slight cross-bedding, which, if it occurs, will slope up in the downstream direction; again, the opposite direction to that seen in ripples and dunes.
At greater velocities still, the current is strong enough to carry the sand in suspension, moving it downstream, leaving only gravel, cobbles, or just plain bedrock, depending on what other sediments, if any, are present on the river bed.
As I have indicated, the type of sediments involved affect these processes: in fine sediments, which "flow" more easily, dunes will not be formed; in coarser sediments, especially in shallower water, the formation of an upper plane bed is less likely, and the sequence as velocity increases will skip straight from dunes to antidunes.
Vanished rivers: how do we know?
Geologists can reconstruct the courses of long-vanished rivers. The method by which they identify them should be obvious and familiar to anyone who has read this far in the textbook. If we take away a river, we are left with its sediments, which will eventually lithify. This will leave us with a set of rocks which look just like the lithified sediments of a river. As usual, we apply the rule that "if it looks like a duck and it quacks like a duck, it's a duck". In the case of rivers there are some very clear indications in the remaining sediment that allow us to identify what it once was.
In the first place, the sediments will be arranged in the long thin form of a river (what is sometimes called a shoestring topography). Note that since rivers shift, the "shoestring" will not necessarily be as narrow as the river was; but it will still be a shoestring. Depending on the depth of river and the rate of flow, the river bed will remain as a shoestring of gravel or cobbles or of duned or rippled sand. The ripples, of course, will cut perpendicular to the direction of the shoestring, making them distinct from beach ripples, which would be parallel to it. As the ripples in rivers are not upstream-downstream symmetric, it also is possible to use them to determine the direction of flow.
The photograph to the right shows a particularly nice example of shoestring topography. In this case, the sedimentary rocks formed from the sediment of a former river have proved more resistant to erosion than the rocks of its former floodplain, and so the former river stands clear as a ridge of sedimentary rock of different composition from that of the surrounding landscape.
Looking horizontally at the different sedimentary types in a line cutting across the direction of flow, we will see coarse sand or gravel at the middle, then finer sand representing point bars, and then the mud of the flood plain. Because rivers gradually shift their course and their banks, we will also be able to see exactly the same sequence going vertically as over time river bed is replaced with point bar is replaced with flood plain. This is known as a fining-up sequence.
The types of sediment will be consistent with the hypothesis of a river. For example, we will not find gypsum or halite, because these would require totally different depositional environments.
Such fossils as we find will be of freshwater plants and animals, or of land plants and animals, but not marine forms. Similarly on the banks of the river we expect such fossils as are present to be of land animals or their footprints, and of land plants.
In short, we see exactly what we should expect to see as the remains of a former river, and are left with the reasonable conclusion that these features are, in fact, the results of the action of a river now vanished.
Note on superposed and antecedent rivers
In many places in the world, we can find rivers which have cut channels through hills or mountain ranges. This may seem odd at first, since naively one might think that since rivers can't flow uphill, they could never have cut the gorges through which they flow in the first place. However, the thing is quite practicable so long as the rivers were there before the mountains.
In the case of antecedent rivers, tectonic uplift slowly raises hills or mountains across the path of the river. So long as the river can erode away the uplifted rock and soil as fast as the rate of uplift, it will maintain its course. As rates of uplift are small compared to the erosional powers of rivers, this should present no problem to a river of reasonable velocity.
In the case of superposed rivers (also known as superimposed rivers) a river flows over a plain, subject to weathering and erosion, beneath which are upward folds (antisynclines) of rock more resistant to erosion than the overlying rock. As erosional processes reveal the resistant antisynclines, the river cuts its way through them, resulting in a river that cuts through hills consisting of the exposed resistant antisynclines. An aerial view of the Susquehanna River, Pennsylvania, can be seen here; the exposed ridges of resistant rock are clearly visible.