Chapter 81

Superimposed Stream Hypothesis—An Act of Desperation

Geologists have commonly switched to the superimposed (sometimes referred to as

superposed) stream hypothesis after rejecting the antecedent stream hypothesis. (A stream here
also refers to a river.) Does this second hypothesis fare any better?

What Is the Superimposed Stream Hypothesis?

A superimposed stream or river is defined as: “A stream that was established on a new
surface and that maintained its course despite different lithologies and structures encountered as
it eroded downward into the underlying rocks.” 1 In this hypothesis, a landscape is buried by
renewed sedimentation, usually by a marine transgression. Then the area uplifts and a stream or
river is established on the generally flat cover of sediments or sedimentary rocks, called the
“covermass.” The top of the covermass is sometimes a planation surface. As erosion takes place
over millions of years, the stream erodes downward in the same location (Figure 81.1). So, after
millions of years the stream ends up flowing through structural barriers.

Figure 81.1 Block diagram of the superimposed stream hypothesis. The stream maintains its course as
most of the covermass (top layer) is eroded (drawn by Bryan Miller).

At the same time, the rest of the cover mass is eroded or mostly eroded in areas far from the
stream, leaving behind the stream or river flowing through ridges or mountains. Apparently, a
geomorphologist will default to this hypothesis based on the discovery of any remnant of what
he thinks is the covermass or even indirect evidence that the covermass once existed. 2 Erosional
remnants of eroded sedimentary rock can usually be found in mountains, so a covermass can
easily be claimed.
Since the antecedent stream hypothesis was usually the hypothesis rejected, this switch at
least shows how weak the antecedent stream hypothesis was all along, as discussed in the

1
Neuendorf, K.K.E., J.P. Mehl, Jr., and J.A. Jackson, 2005. Glossary of Geology, Fifth Edition. American
Geological Institute, Alexandria, VA, p. 645.
2
Twidale, C.R., 2004. River patterns and their meaning. Earth-Science Reviews 67:194.
previous chapter. So, the superimposed stream hypothesis seems to be favored for explaining the
many examples of anomalous drainage. 3

The Classic Appalachian Mountains Explanation

Discordant drainage in the Appalachian Mountains (see Chapter 77) was first attributed to
antecedent streams, as one might expect since it was the first hypothesis proposed. 4 However,
superimposition is now favored, partly because the ridges are believed to define a planation
surface. 5 The planation surface is generally level, and rivers flowing on this surface were
assumed to have cut down into older deformed sedimentary rocks. The idea is applied mainly to
large rivers; smaller streams are generally congruent with the geological structure, in other words
the tributaries mostly flow down tributary valleys. Ollier noted that the major rivers were
attributed to superimposition while their tributaries usually were not:
Classic examples of superimposed drainage are found in the Appalachian Mountains.
Accordant levels of ridge tops show that the folded Palaeozoic strata were planated, and
major rivers such as the Susquehana [sic] originally flowed across this plain regardless of
structure. They have later been incised and now flow through superimposed gorges, but
tributaries are strongly structurally controlled. 6
However, even some tributaries also cut through barriers in the Appalachian Mountains (see
Chapter 77).Logically, the water flowing through tributaries would be too weak to carve
downward through the covermass.
Von Engeln believed that the remarkably aligned water gaps of the Appalachian Mountains
are good evidence for superimposition, since it is difficult to conceive of aligned water gaps
being caused by antecedence or stream piracy (see the next chapter).4 (A better explanation for
aligned water gaps will be given Part XVIII.) Strahler saw two possibilities for Appalachian
transverse drainage: superimposition and fault control. 7 He left no room for stream piracy, and
claimed that wind gaps are evidence of superimposition. Fault control has already been
eliminated for practically all Appalachian water and wind gaps. 8 Superimpostion seems to have
been accepted mainly because of difficulties with other hypotheses, rather than because of strong
positive evidence.
The hypothesis proposes the Appalachian covermass developed after uplift and deformation.
The mountains sank again during the Cretaceous period of geological time and were
subsequently covered by sedimentary rocks from a marine transgression. Upon renewed uplift,
rivers formed on the horizontal top of the sedimentary rocks, and over millions of years eroded
downward without deviation, eventually cutting through barriers and creating water gaps.

3
Small, R.J., 1978. The Study of Landforms: A Textbook of Geomorphology, second edition, Cambridge University
Press, London, U.K., pp. 238-247.
4
Von Engeln, O.D., 1942. Geomorphology: Systematic and Regional, Macmillan, New York, NY.
5
Short, N.M. and R.W. Blair, Jr. (editors), 1986. Geomorphology from Space: A Global Overview of Regional
Landforms, NASA, Washington, D.C., p. 56.
6
Ollier, C., 1991. Ancient Landforms, Belhaven Press, New York, NY, p. 33.
7
Strahler, A.N., 1945. Hypotheses of stream development in the folded Appalachians of Pennsylvania. GSA Bulletin
56:45-88.
8
Clark, G.M., 1989. Central and southern Appalachian water and wind gap origins: review and new data.
Geomorphology 2:209-232.
 Unfortunately, there is no evidence for the proposed transgression and the great volume of
“covermass.” 9 Furthermore, why wouldn’t the stream change course to flow on the presumably
softer covermass rather than through hard-rock anticlines? As with the antecedent stream
hypothesis, the superimposition idea suffers from a lack of evidence. Mills and others
summarized:
As for superposition from an unconformable cover mass, there is no evidence of such a
cover, and unlike in the ancient Appalachians, it is much more difficult to claim that the
cover mass has been removed by erosion. 10
Ollier also admitted there is no evidence for the Appalachian covermass:
The age of the old planation is controversial, as is the former existence of a Cretaceous
cover. The lack of any remnants of a Cretaceous cover makes the idea questionable.6
Because of the lack of a covermass, Epstein rejected regional superimposition while accepting
“local” superimposition. 11
Some authors claim that the Appalachian rivers first became established on a “planation
surface,” but a covermass is a depositional surface not a planation surface, unless they believe
the covermass consolidated and was planed off by water, somehow.

Southern England Examples

Superimposition has been used to explain the discordant drainage in Wales. 12 That
interpretation is more an inference or interpretation, because it assumes the presence of a
Cretaceous chalk deposit laid down during a marine transgression and subsequently eroded.
Chalk covers much of southeast England (Figure 81.2). Therefore, it is simply assumed that the
streams were first superimposed on a level chalk surface. Geologists infer a widespread chalk
substrate from erosional remnants. At least in this case, there is a remnant of the “covermass,”
unlike the Appalachians.
However, these remnants do not necessarily mean that the strata were either continuous or
horizontal when first deposited, even though both are commonly assumed. Nor does it mean that
streams followed the same course for millions of years, eroding both chalk and then causing
water gaps through more resistant rock. There actually is no evidence for the superimposition of
the Wales drainage. R.J. Small, although believing in the mechanism, nevertheless admitted:
“However, superimposition from a Chalk cover is not the only available explanation of such
discordances between streams and structures.” 13
Southeast England also has been considered a classic case of superimposition for the water
gaps located there, but this interpretation is also easily questioned. 14

9
Mills, H.H., G.R. Brakenridge, R.B. Jacobson, W.L. Newell, M.J. Pavich, and J.S. Pomeroy, 1987. Appalachian
mountains and plateaus, In, Graf, W.L. (editor), Geomorphic Systems of North America, Geological Society of
America Centennial Special Volume 2, Boulder, CO, p. 13.
10
Mills et al., Ref. 9, p. 14.
11
Epstein, J.B., 1966. Structural control of wind gaps and water gaps and of stream capture in the Stroudsburg area,
Pennsylvania and New Jersey. U.S. Geological Survey Professional Paper 550-B, Washington, D.C.
12
Small, R.J., 1978. The Study of Landforms: A Textbook of Geomorphology, second edition, Cambridge University
Press, London, U.K., pp. 243-247.
13
Small, Ref. 12, p. 247.
14
Chorley, R.J., S.A. Schumm, and D.E. Sugden, 1984. Geomorphology, Methuen, London, U.K., p. 21.
 Figure 81.2. White Cliffs of Dover, southeast England (Wikipedia).

Little or No Evidence for Superimposition

Besides the problems mentioned in the two examples above, other questions remain on the
validity of this hypothesis.
Mabbutt rejected both the antecedent and superimposition hypotheses for the water gaps
through the Macdonnell Ranges of central Australia 15 After ruling out stream piracy and
structural weaknesses, he concluded the transverse drainage was “inherited” directly from an
erosion surface but without superimposition because the main water gaps are aligned. This
explanation currently is not even considered as one of the hypotheses for the origin of the water
gaps. Regardless, he eliminates all the other suggested hypotheses.
Although geologists at first believed that the Rocky Mountain water gaps were formed by
antecedent streams, later most embraced the superimposed stream hypothesis. But Hunt was still
skeptical because the valley fill would have to be up to the tops of the mountains:
However, the stream courses across the various ranges in the Rocky Mountains probably
are not superimposed. Too much fill would have been required to bury the several
mountain ranges, and too much erosion would have been required to remove that fill. 16
One of the main problems with the superimposition hypothesis is the river must maintain the
same course and downcut into both resistant and non-resistant formations, while its tributaries,
along with mass wasting, erode the remainder of the covermass. 17 C.H. Crickmay was skeptical
of the entire process, since erosion would be weak away from rivers and streams:
Superimposition, on the other hand, demands two distinct requirements: the stream must
maintain, by down-cutting, a course across a resistant formation into which, along with
all the other formations, it is incising itself; at the same time, it must swing to-and-fro
across all the remainder of the country (all its tributaries, presumably, doing the same)
both upstream and down from the resistant zone, and must remove enough of the less
resistant material to make the more resistant stand out in relief … For some reason not

15
Mabbutt, J.A., 1966. Landforms of the Western Macdonnell Ranges. In, Dury, G.H. (editor), Essays in
Geomorphology, Heinemann, London, U.K., pp. 112-114.
16
Hunt, C.B., 1967. Physiography of the United States, W.H. Freeman and Company, San Francisco, CA, p. 272.
17
Douglass, J. and M. Schmeeckle, 2007. Analogue modeling of transverse drainage mechanisms. Geomorphology
84:40–41.
 too evident, no textbook has gone far enough in its thinking on superimposition to touch
on this requirement, much less, to conceive of a mode of meeting it (emphasis his). 18
Since there is scant or no evidence of superimposition it leaves the idea without validation:
Although a plausible mechanism, superimposition is extremely difficult to verify except
in the case of very young orogens [uplifted linear, folded, and deformed mountain belts]
where vestiges of the original sedimentary cover remain. In ancient mountain belts,
denudation will have removed all the evidence of any pre-existing sedimentary cover. 19
Even if a remnant of a sedimentary formation that would qualify as a patch of covermass, it
still has to be demonstrated that the strata were once continuous and horizontal, and that erosion
occurred while the major streams remained locked in place. Since most of the strata have been
eroded, it is in fact an argument from a lack of evidence—typically one of the weakest types.
Table 81.1 summarizes the evidence against the hypothesis.

1. Lack of evidence for a transgression of the sea and/or a covermass

2. Most, if not all, covermass eroded from area while rivers concentrate erosion linearly
3. Usually no evidence
4. Erosional remnants do not prove a covermass
5. In some cases, the covermass volume is huge and erosion must be great
6. Change in geological structure or lithology does not deflect the stream
7. Stream must maintain the same course even after softer valley rocks eroded
Table 81.1. Problems with the superimposed stream hypothesis.

18
Crickmay, C.H., 1974. The Work of the River: A Critical Study of the Central Aspects of Geomorphology,
American Elsevier Publishing Co., New York, NY, p. 155.
19
Summerfield, M.A. 1991. Global Geomorphology, Longman Scientific & Technical, New York, NY, p. 411.