ArticlePDF Available

Erosion, isostatic response, and the missing peneplains


Abstract and Figures

The peneplain—a low-relief erosional plain worn to near base level—is a venerable concept in geomorphology, geology, and geography. Yet despite more than a century of effort, no convincing example of a contemporary peneplain has been identified, and the identification of relict peneplains is uncertain and controversial. As a peneplain is a logical outcome during a period of long tectonic stability, the paucity or absence of peneplains is problematic. Most explanations are based on the notion that the periods of tectonic stability required for peneplain formation are too long to allow the features to fully develop, or that Neogene tectonics has precluded recent peneplanation. This paper proposes an alternative explanation, generally consistent with those given above, which can also explain the absence of peneplains in regions experiencing long tectonic stability. If erosion or deposition rates are related to elevation and if isostatic response (uplift or subsidence) is related to erosional unloading or depositional loading, the relationship between these components is dynamically unstable. This is demonstrated mathematically. This instability implies that no particular state or mode of topographic evolution, including peneplanation, is likely to persist in the face of variations or perturbations that influence any system component. Thus, formation of a peneplain would require tectonic stability and also relative constancy in sea level (or rates and direction of sea level change), climate, biotic influences on erosion or deposition, and any other factors that modify erosion, deposition, elevation fields, or isostatic responses. This would explain an absence of geologically contemporary peneplains and a rarity of well-developed peneplains in the geologic record.
Content may be subject to copyright.
Erosion, isostatic response, and the missing peneplains
Jonathan D. Phillips *
Tobacco Road Research Team, Department of Geography, University of Kentucky, Lexington, KY 40506, USA
Received 11 June 2001; received in revised form 28 September 2001; accepted 4 October 2001
The peneplain—a low-relief erosional plain worn to near base level—is a venerable concept in geomorphology, geology,
and geography. Yet despite more than a century of effort, no convincing example of a contemporary peneplain has been
identified, and the identification of relict peneplains is uncertain and controversial. As a peneplain is a logical outcome during a
period of long tectonic stability, the paucity or absence of peneplains is problematic. Most explanations are based on the notion
that the periods of tectonic stability required for peneplain formation are too long to allow the features to fully develop, or that
Neogene tectonics has precluded recent peneplanation. This paper proposes an alternative explanation, generally consistent with
those given above, which can also explain the absence of peneplains in regions experiencing long tectonic stability. If erosion or
deposition rates are related to elevation and if isostatic response (uplift or subsidence) is related to erosional unloading or
depositional loading, the relationship between these components is dynamically unstable. This is demonstrated mathematically.
This instability implies that no particular state or mode of topographic evolution, including peneplanation, is likely to persist in
the face of variations or perturbations that influence any system component. Thus, formation of a peneplain would require
tectonic stability and also relative constancy in sea level (or rates and direction of sea level change), climate, biotic influences on
erosion or deposition, and any other factors that modify erosion, deposition, elevation fields, or isostatic responses. This would
explain an absence of geologically contemporary peneplains and a rarity of well-developed peneplains in the geologic record.
D2002 Elsevier Science B.V. All rights reserved.
Keywords: Peneplain; Peneplanation; Erosional unloading; Isostatic response; Topographic evolution; Landscape evolution
1. Introduction
The peneplain—a low-relief erosional plain worn
to near base level—is a venerable concept in geo-
morphology, geology, and geography. Cyclic theories
of landform development, particularly those deriving
from the work of William Morris Davis, typically
postulate that denudation should lead to a peneplain.
The search for ancient and modern peneplains has
thus been a major theme of cyclic geomorphology
(see Chorley et al., 1973; Beckinsale and Chorley,
1991 for a review and historical perspective).
While some geomorphologists, not inappropriately,
associate peneplains with Davisian cyclic geomor-
phology (cf. Ritter, 1988), the term and the concept
are still very much in use. Cyclical geomorphology
has declined in popularity, but contemporary studies
of landscape evolution, tectonic geomorphology, and
historical geology routinely use the term peneplain;
and interpretations of landform evolution, tectonic
history, and lithological controls (among other thing)
0169-555X/02/$ - see front matter D2002 Elsevier Science B.V. All rights reserved.
PII: S0169-555X(01)00156-8
Tel.: +1-859-257-6950; fax: +1-859-323-1969.
E-mail address: (J.D. Phillips).
Geomorphology 45 (2002) 225 241
are often based on the presence of low-relief erosion
As the search for peneplains has gone on for more
than a century, strikingly few, if any, contemporary
peneplains have been identified. Various kinds of
erosion surfaces exist, and evidence (sometimes dis-
puted) is available of remnant peneplains at elevations
well above Holocene sea levels. However, extensive
low-relief erosional surfaces worn to near sea level are
at best rare, and by some reckonings nonexistent. This
question is important well beyond cyclic theories of
landscape evolution. Summerfield (1996, p. 212)
holds that ‘‘there is arguably no more fundamental
data pertinent to the understanding of long-term land-
scape development than alterations in surface eleva-
tion through time since these actually define changes
observed in landscape morphology with respect to the
‘absolute’ datum of the geoid.’’ Thus, relationships
between elevation, uplift, and denudation are critical
for geomorphology and geophysics.
This paper examines the question of why few (or
no) modern peneplains exist, in the context of inter-
actions between erosional unloading (or depositional
loading) and isostatic response. I will be argue that the
paucity of peneplains can be explained on the basis of
the fundamental nature of the relationship between
erosion, deposition, elevation, and isostatic adjust-
2. Peneplains and peneplanation
An exhaustive history of peneplain concepts is
beyond the scope of this review and is unnecessary
in light of the excellent geomorphological histories by
Chorley et al. (1973) (focussing on the life and work of
W.M. Davis) and Beckinsale and Chorley (1991)
(covering historical geomorphology up to about 1950).
At the risk of yet another oversimplification of
Davis’ ideas, cycles of erosion were postulated to lead
inexorably (in the absence of renewed uplift) toward a
low-relief surface dominated by denudational removal
and worn to near base level or sea level. Because
some relief must be present and truly flat erosional
surfaces would defy geophysical principles, Davis
coined the term peneplain, or ‘‘almost-plain.’’ Davis’
(1898) study of the Triassic basins of New Jersey is a
well-known exemplar of his application of the pene-
plain concept to geological explanation. An 1899
article titled simply ‘‘The Peneplain’’ summarizes
Davis’ views and includes a defense against attacks
on the concept and its application in the eastern U.S.
(Davis, 1899). A later detailed exposition on pene-
plains and the cyclic model of landscape evolution
(Davis, 1922) addresses some subsequent criticisms
and incorporates some adaptations of the concept. In
Davis (1932), the peneplain concept is again
defended, and contrasted with the ideas of Penck
Davis’ definition of peneplains was unquestionably
flexible, as evidenced by (among other things) his
adaptation to arid lands (Davis, 1905). Considerable
latitude was allowed in the degree of downwearing,
for example. A diagram of a peneplain from Davis
(1912) shows monadnocks (resistant residual hills),
rolling hills, significant dissection, and a relief of 700
ft even at the lowest-relief margin. Whether this
flexibility is viewed as inherently flawed and impre-
cise (cf. Ritter, 1988) or as merely clumsily applied by
those less talented than Davis (see Chorley et al.,
1973), the definitional looseness meant that ‘‘every
geologist was free to establish his own meaning for
the term,’’ with many earth scientists conflating any
erosion surface with a peneplain (Ritter, 1988, p. 3).
Sparks (1960, p. 334) distinguishes different types
of erosion surfaces, but attaches cyclic significance to
certain types:
Strictly speaking, any surface which is not an
original structural or constructional surface is an
erosion surface... But, in practice, it is applied
essentially to a surface of faint relief, the end
products of complete or incomplete cycles of
Peneplains and other erosion surfaces, such as
pediplains, correspond to a loose definition of ‘‘a
landscape of low relief resulting from prolonged
subaerial erosion,’’ although the processes creating
them may differ significantly (King, 1953, pp. 738
739). A recent discussion distinguishes peneplains
from other types of planation surfaces, such as pedi-
ments and etchplains (Cui et al., 1999).
A peneplain should fulfill the following conditions:
(i) a dominantly erosional surface, with erosion pri-
marily fluvial and subaerial; (ii) cut to near base level;
J.D. Phillips / Geomorphology 45 (2002) 225–241226
(iii) low relief; (iv) truncating all rock types of differing
resistance; and (v) subcontinental in extent. To be
considered a Davisian peneplain, a sixth condition
should be added, that the surface represents the ultimate
stage of a cycle. The term peneplain is, however, well
entrenched in the literature and is routinely applied to a
variety of low-relief erosion surfaces where subaerial
processes have worn the surface to near base level (or
where this is believed to have occurred). The remainder
of this paper will accept a broad definition of pene-
plains that meets only the first five conditions above.
2.1. Peneplains in recent literature
Ritter (1988) maintains that referring to an erosion
surface as a peneplain implies adherence to cyclic
concepts. Recent literature on landscape evolution and
using morphology and erosion surfaces to deduce
tectonic settings and histories, however, makes free
and frequent use of the term. Direct reference to
Davisian cycles is rare, and in general, assumptions
about the role of peneplains or erosion surfaces in
geological cycles are unnecessary to the arguments
and interpretations. The role of subsequently uplifted
and dissected planation surfaces is critical in the
theory of mountain-building proposed by Ollier and
Pain (2000), for instance. However, while they rec-
ognize peneplains as one type of planation surface,
planation surfaces themselves are critical to the argu-
ment, not the particular mode or mechanism of for-
mation (Ollier and Pain, 2000). They use the term
peneplain in its most general meaning, as an ‘‘almost-
Much recent use of peneplain concepts and termi-
nology is in the context of interpreting crustal move-
ments. Higgins (1975) noted that interest in finding
ancient erosion surfaces was renewed with the advent
of plate tectonic theory, and he identified a transition
from a search for peneplains as evidence of geo-
morphic cycles to a search for peneplains as potential
indicators of uplift histories. Coltori and Ollier (2000),
for example, made use of a lower Pliocene planation
surface graded to sea level in their interpretation of the
geomorphic and tectonic evolution of the Ecuadoran
Andes. A Cenozoic peneplain surface preserved in an
extensional wedge of the western Alps is crucial in
Eder and Neubauer’s (2000) interpretation of Neogene
morphotectonics. The Central Otago peneplain, New
Zealand, provides evidence of tectonic history (Stir-
ling, 1990; Markley and Norris, 1999) and of erosion
history since the Miocene (Stirling, 1991). Landforms,
including residual erosion surfaces and an exhumed
peneplain, formed the basin for Lidmar-Bergstrom’s
(1999) deduction of uplift history of Scandinavian
In the Scandinavian region, peneplains are crucial
to recent interpretations of physiography, geologic
history, and geomorphic evolution. Riis (1996) argued
that the correlation of offshore geology and onshore
morphology suggests that the enveloping summit
level of Scandinavia originated as a Jurassic pene-
plain. Tikkanen (1994) recognized much of Finland as
a peneplain, but with till and other covers. Most of the
Baltic shield was exposed in the late Proterozoic,
leading to an extremely flat primary peneplain. This
surface was subsequently covered by Paleozoic rocks,
thus preserving the surface until re-exposure (Lidmar-
Bergstrom, 1995, 1999). The Pre-Cambrian basement
in this region is a mosaic of exhumed surfaces,
including a sub-Cambrian peneplain, and new surfa-
ces formed after erosion of cover rocks. Lidmar-
Bergstrom (1996) recognizes the sub-Cambrian pene-
plain first identified by earlier workers and observes
that the relief of the cratons was difficult to fit into a
system of erosion cycles. Johansson (1999) recog-
nized peneplains among palaeosurfaces in SW Swe-
den; a ‘‘very even’’ sub-Cambrian peneplain in the
east and an ‘‘uplifted and broken’’ part of the pene-
plain in central portion. Johansson et al. (1999) used
digital elevation and geologic data with geostatistical
analysis to describe the morphotectonics of the frag-
mented part of the sub-Cambrian peneplain. The
feature is separated into fault-bounded blocks. Their
analysis revealed a varied relief different from the
extremely flat Precambrian surface usually described
near Cambro Silurian cover rocks in southern Swe-
Peneplains also figure into general descriptive, the-
oretical, and methodological works in the contempo-
rary geomorphology and geology literature. The atlas
of landforms produced by Blume (1992) includes three
examples of peneplains where the term is used explic-
itly and two other references to erosion surfaces. Cui et
al. (1999) presented a general discussion of planation
surfaces and the geologic interpretations of such sur-
faces, distinguishing between peneplains, pediments,
J.D. Phillips / Geomorphology 45 (2002) 225–241 227
and etchplains. The distinction is based mainly on
whether fluvial downwearing, slope backwasting, and
weathering is the primary denudational process. A
Markov model approach to relief evolution based on
transition probabilities was developed by Gournellos
(1997), who identified 16 states or stages of relief
development in an assumed evolutionary sequence
(based on the inability to move between some stages
without passing through intervening stages). Depend-
ing on the relative magnitudes of erosion, uplift, and
subsidence over time, many possible landform states
can develop. The peneplain is recognized as an
‘‘absorbing state’’ achieved when the whole available
relief is eroded and degradation stops.
The discussion above is illustrative, not exhaustive,
and is intended to show that the peneplain concept is
alive and well in earth science, and not necessarily
tied to any particular theory of landscape evolution.
While the arguments are often dependent on a plana-
tion surface, they are typically independent of whether
the surface is a peneplain, pediplane, etchplain, or
marine planation surface.
3. Where are the peneplains?
Although earth scientists have searched for pene-
plains in one context or another for more than a century,
modern peneplains are almost universally recognized
as scarce or even nonexistent. While relict or remnant
low-relief subaerial erosion surfaces believed to have
been cut to near base level are widely recognized, a
contemporary surface eroded to near modern sea level
has yet to be convincingly identified (recognizing that
local and regional base levels may not correspond with
sea level).
The paucity of peneplains has long been acknowl-
edged. Sparks (1960) noted that good examples of
peneplains are rare and acknowledges that some earth
scientists would say they do not exist. Ritter (1988, p.
4) held that ‘‘...peneplains are indeed a rare com-
modity if, in fact, they exist or ever existed.’’ While
‘‘there is no question that widespread erosion surfaces
exist,’’ most are not peneplains (Ritter, 1988, p. 4).
Even stronger is King’s (1953, p. 749) assertion: ‘‘A
peneplain in the Davisian sense, resulting from slope
reduction and downwearing, does not exist in nature.
It should be redefined as ‘an imaginary landform.’’
King (1953) does, however, recognize the existence of
pediplains, which are morphologically similar to pene-
plains (though pediments, the major element of ped-
iplains, are typically concave upward in contrast with
the convex crests of topographic highs in a rolling
peneplain). Indeed, W.M. Davis always admitted to the
rarity or absence of geologically recent peneplains,
attributing this to crustal upheaval and asserting that
peneplains were more numerous in the geologic past.
Davis (1899, pp. 232 233) acknowledged that ‘‘there
are today no extensive peneplains still standing close to
the sea level with respect to which they were denuded,’’
but argued that fossil examples provide abundant
evidence of their occurrence (p. 232 –233). ‘‘Admitting
the present to be exceptional in the lack of peneplains
close to their base level of production,’’ he postulates
‘‘general disturbances by uplift and tilting in the recent
past’’ (Davis, 1899, p. 223).
Even the staunchest adherents of peneplanation
were obliged to recognize the absence of contempo-
rary low-relief subaerial erosion surfaces denuded to
near sea level or other contemporary base levels.
However, the recognition of relic peneplains and the
interpretation of erosion surfaces as peneplains are also
controversial. An early critic of the peneplain concept
was Tarr (1898), who was concerned with the paucity
of recent erosion surfaces conforming to the penepla-
nation concept, and also challenged evidence of a relict
peneplain in New Jersey and New England. Tarr
(1898) maintained that Davis’ arguments in this regard
were based on weak evidence and presented topo-
graphic data to show that the accordancy of summit
elevations, critical to interpretations of an ancient
peneplain, did not exist. The Tarr (1898) vs. Davis
(1898, 1899) peneplain debate was discussed by
Rhoads and Thorn (1996) as an example of the effects
of theory-ladenness in geomorphology. They generally
supported Tarr’s arguments, suggesting that a relic
peneplain may only be detectable if one is already
conditioned to believe that such a feature is likely to
exist. The interpretation of ancient Appalachian ero-
sion surfaces and the recognition of uplifted and
dissected peneplains in this region has been subject
to dispute for the better part of a century (see review in
Beckinsale and Chorley, 1991, Chap. 7).
Interpretation of erosion surfaces and recognition
of peneplains in the topographic and geologic record
has also been controversial elsewhere. Identification
J.D. Phillips / Geomorphology 45 (2002) 225–241228
of a planation surface is not always straightforward.
Once that is accomplished, attributing its origin to
fluvial downwasting, pediplanation, etching, marine
planation, or other mechanisms such as the combina-
tion of duricrust formation and subsurface flushing
proposed by Trendall (1962), is not easy. The notion
of a Tertiary or earlier ‘‘great Australian peneplain,’
for example, is disputed by Twidale and Campbell
(1995) on grounds that though there was general
Tertiary planation, there were distinct regional varia-
tions in processes and timing. The Australian con-
tinent has abundant evidence of Quaternary planation
of extensive areas, with rolling surfaces of low relief
well represented. Many of these surfaces ‘‘which
morphologically resemble peneplains’’ result from
fluvial erosion, but others lack surface drainage and
may be attributed to differential subsurface weath-
ering and flushing (Twidale and Campbell, 1995, pp.
1920). They argued more generally that the great
antiquity of many Australian landforms and land-
scapes made them inconsistent with cyclic theories
of landscape evolution. On the other hand, a looser
and more general definition of a peneplain as ‘‘a
rolling or undulating surface of low relief probably
resulting from lowering of the surface’’ was deemed a
useful concept in landscape description by Twidale
and Bourne (1998, p. 113), even as they acknowledge
that Davisian peneplanation has not yet been demon-
strated. Low-relief palaeoplains were recognized by
Ollier (1995) in SE Australia, but evolution through a
stage of plateau remnants to knife-edge ridges is ‘‘the
reverse of ‘peneplanation’ as commonly conceived’’
(p. 43). In the Front Range of the western U.S.,
paleobotanical evidence suggests the erosion surface
capping the front range was at 2.2 to 2.3 km elevation,
not near base level as a peneplain origin would require
(Gregory and Chase, 1994).
The recognition of ancient, uplifted peneplains
hinges in large part on the accordancy of plateau
surface, drainage divide, or summit elevations and
the interpretation of those similar elevations as having
once been denuded close to a single base level.
Equifinality is a critical problem because a number
of other plausible explanations exist for accordancy of
summits and divides.
First, even if topographic accordances represent an
uplifted low-relief erosion surface, such a surface
could have been produced by mechanisms other than
peneplanation, including pediplanation, marine plana-
tion, etchplanation, or lateral stream erosion (Penck,
1924; King, 1953; Palmquist, 1975; Beckinsale and
Chorley, 1991, pp. 235– 237; Ollier, 1995; Cui et al.,
1999). Second, topographic accordances may be
structurally, lithologically, tectonically, or isostatically
controlled (Palmquist, 1975; Beckinsale and Chorley,
1991, pp. 235 237; Gregory and Chase, 1994; Mont-
gomery, 1994). Third, various arguments exist that
topographic accordance may be created by dynamic
adjustments to common geological and environmental
controls (Hack, 1960; Beckinsale and Chorley, 1991,
pp. 235237; Dykes and Thornes, 1996). Fourth,
regularity of stream spacing may produce accordance
of divides as downcutting is balanced by isostatic
adjustments (the Gipfelflur hypothesis; Penck, 1924;
Dykes and Thornes, 1996). A fifth possibility is raised
by Harbor (1997), who showed that during basin
expansion in the northern U.S. basin and range,
stream erosion exceeded hillslope retreat, elevating
summit plateaus.
A sixth possible mechanism, devised by Pavich
(1989), explains the flatness of the Appalachian pied-
mont drainage divides, which provide the appearance
of a planation surface of accordant summits. Pavich
argued that the topography resulted from a combina-
tion of three active processes: volume reduction dur-
ing compaction of saprolite into soil, loss of mass in
dissolved solids draining the soil B-horizon, and loss
of mass from the soil surface by erosion of clay-sized
particles. The cumulative effect is to convert saprolite
to soil, to limit soil thickness and residence time, and
to limit the age of the oldest residual soils on the
surface. This results in a continual lowering of the
‘‘peneplain’ surface (Pavich, 1989). Gregory and
Chase (1994) show, based on work in the Front
Range, that regional erosion surfaces may represent
regional climatic rather than tectonic conditions.
While the high elevation of the summits results from
tectonic forces, models suggest the smoothness is a
function of climate, identifying a seventh general
origin for accordant summits.
At least seven plausible explanations for topo-
graphic accordances exist. These explanations indi-
cate that not only are peneplains rare or absent as
contemporary landscape features, but that they are
also difficult to recognize as relics or remnants and
that ancient peneplains may be less common than
J.D. Phillips / Geomorphology 45 (2002) 225–241 229
some have assumed or asserted. Yet, the fundamental
reasoning behind the peneplain concept is sound. In a
tectonically stable landscape dominated by subaerial
fluvial erosion, a progressive lowering of relief, con-
trolled by base level, is expected and is predicted by
physically based models. As Palmquist (1975, pp.
161 162) puts it: ‘‘Conceptually, no reason exists
why, during a sufficiently long period of tectonic
stability, the landforms of a stable region should not
evolve to produce the surface of low relief known as a
peneplain.’’ A proper question is, therefore, ‘‘where
have all the peneplains gone?’’
3.1. Where have all the peneplains gone?
One answer, based on the discussion just above,
is that there never were any or many peneplains and
that the real and apparent erosion surfaces evident in
the topographic and geologic record were produced
by other processes and mechanisms. Another explan-
ation, favored by Davis (1899), among others, is that
the Quaternary has been too tectonically unstable to
facilitate the development of peneplains (Sparks,
1960; Palmquist, 1975). Generally, several arguments
have been made that peneplanation requires unreal-
istically long periods of crustal quiescence, and that
this accounts the absence of peneplains. Pitman and
Golovchenko (1991), using a model coupling ero-
sion, isostatic response, elevation, and base level
change, suggested that peneplanation can only occur
during periods of sustained sea level rise on the
order of 250 m or more over 50 million years or
The notion of global trends in tectonic activity is
controversial but plausible. Ollier and Pain (2000)
presented evidence of a roughly synchronous global
episode of Plio Pleistocence uplift, and also suggest
that global periods of ‘‘tectonic stillstand’’ must have
occurred to allow the formation of numerous exten-
sive planation surfaces. However, as King (1967)
eloquently stated:
The planet is a structural entity and we shall
rightly anticipate that major stresses may be
relieved, and major orogenic events occur, with a
broad simultaneity over its surface. But the
planetary crust is a thing of infinite heterogeneity
in detail, and we shall also anticipate that local
deviations from orogenic similarity and simulta-
neity will everywhere be the rule (p. 429).
Thus, even during a time of generally high tectonic
activity, we should expect some areas of relative
stability. More to the point, one objection to the lack
of tectonic stability as the explanation for the absence
of peneplains is that significant portions of the earth
surface have experienced quite long periods of tec-
tonic stability. Australia and other Gondwanan land
surfaces, in particular, show evidence of long periods
of crustal stability and ancient landforms (Twidale,
1994, 1999a,b; Paton et al., 1995), but no peneplains.
This paper will advance the argument that pene-
planation represents a particular state or mode of
landscape evolution. The fundamental relationships
between erosional unloading (or depositional load-
ing), isostatic uplift (or subsidence), and elevation are
such that the landscape system is fundamentally
unstable. This means that any change to one or more
system components (for example, associated with
tectonic, climatic, or biological changes) is likely to
persist and grow over finite time, meaning that any
given mode or state of landscape development
(including peneplanation) is unlikely to persist. This
fundamental instability in the face of environmental
change accounts for the absence of peneplains or other
evidence of topography requiring very long periods of
dominance by a specific mode or state of landscape
evolution. This argument will be developed below.
4. Erosion isostasy elevation interactions
4.1. States of landscape evolution
The state of a landscape or geomorphic system
may be defined as a particular qualitative condition or
mode of operation. A hillslope, for example, may be
in a state dominated by fluvial erosion. While rates of
erosion and other relevant processes may vary quan-
titatively, as long as removal of material from the
hillslope by wash erosion exceeds production of new
material by weathering and input from deposition, the
state of the system remains constant. An often-useful
approach to landscapes and geomorphic systems is to
examine them in terms of the transitions between
system states. A good example of this approach is
J.D. Phillips / Geomorphology 45 (2002) 225–241230
Kocurek and Lancaster’s (1999) study of transitions in
the aeolian system state, with the states defined by the
relationships between sediment supply, sediment
availability for deflation, and the transport capacity
of wind. In this study, system states are associated
with the relationships between erosion/deposition,
isostatic compensation, and elevation.
In the context of topographic evolution, Gournellos
(1997) identified 16 states or stages of relief develop-
ment based on the relative probabilities of uplift move-
ments and erosion. In addition, transitions between
states are constrained by evolutionary sequences, e.g.,
the relief cannot move between some states without
passing through intervening states. Gournellos’ proba-
bilistic, Markov-based approach identified a peneplain
as an attractor or absorbing state achieved when the
whole relief is eroded and identified many possible
landform states (rather than any particular cyclic or
evolutionary sequence) that may develop. This
approach suggests that development of a peneplain is
at least improbable in two senses. First, it is only one of
many possible states. Second, peneplanation in the
Gournellos (1997) model depends on the unlikely
prospect of complete denudation of relief and reaching
a static condition of uplift and erosion.
A second approach is that of Kennedy (1962), who
identified nine different cases of landscape develop-
ment based on the relative rates of tectonic uplift,
stream downcutting, and slope denudation. Kennedy
(1962) did not account for the effects of deposition, and
assumed that all slope denudation reduced slope gra-
dients and elevations. A more general version of this
approach is that of Phillips (1995a), who considered
topographic development in the context of whether
randomly selected pairs of locations tend, on average,
to converge or diverge in elevation. Ten different
system states were distinguished, based on whether
initially higher and initially lower points are increasing,
decreasing, or remaining constant in elevation. Five of
the states are stable, characterized by convergent ele-
vation (relief is reduced over time). The other five are
unstable, characterized by divergent evolution where
elevation differences, on average, increase over time.
While the nonlinear dynamical systems approach to the
evolution of relief in Phillips (1995a) differed from
earlier studies, the paper showed that a number of
existing theories, models, and conceptual frameworks
for landscape evolution recognize modes of evolution
involving both increasing and decreasing relief ampli-
tude. Isostatic responses to erosion and deposition
were not explicitly considered by Phillips (1995a),
who did not distinguish between isostatic and other
forms of tectonic uplift or subsidence. Kennedy (1962)
did not discuss isostasy, and stated outright that there is
‘‘no direct relationship between denudation and tec-
tonic movement’’ (p. 306).
If elevation changes at initially higher or lower
points are divided into those attributable primarily to
erosion/accretion and those from uplift/subsidence, 18
possible states occur (Table 1).
With respect to production of a peneplain, relief
evolution modes 10 18 in Table 1 all involve diverg-
ing elevation and increasing relief and cannot produce
a planation surface. Modes 3, 5, and 7 (also 13, 14, and
17) call for variable rates of uplift. This is certainly
plausible over very large areas or at a local scale (for
instance, either side of a fault zone). Over intermediate
sizes and scales, in the context of average responses of
pairs of locations, this variable uplift is unlikely.
Models of the interaction of uplift, erosion, and ele-
vation or relief are based on uniform rates of uplift
within the modelled area (Ahnert, 1970, 1984; Lam-
beck and Stephenson, 1986; Pinet and Sourian, 1988;
Pitman and Golovchenko, 1991; Bishop and Brown,
1992; Masek et al., 1994; Montgomery, 1994; Zhou
and Stu
¨ne, 1994; Dykes and Thornes, 1996). Modes
2 and 6 (also 12 and 18) apply to depositional sit-
uations; and mode 9, postulating a uniform surface
with complete isotropy in erosion/deposition and
uplift, is completely unrealistic and useful only as an
This leaves 3 of 18 modes (1, 4, and 8 from Table 1)
that might produce a peneplain. Mode 1 is erosional
and requires that higher sites erode at a faster rate than
lower sites (otherwise, relief will increase or remain
constant). Mode 4 calls for erosion of higher and
accretion of initially lower locations. This could lead
to a peneplain if erosion exceeds accretion so that
denudation occurs. Mode 8 calls for erosion of initially
higher and no elevation change at lower sites. These
three modes have in common a net removal of mass
from the landscape. If this denudational unloading
persists long enough, some isostatic compensation
must occur, potentially moving the system into another
state. The model of nonlinear dynamical systems thus
implies that relief development modes that could pro-
J.D. Phillips / Geomorphology 45 (2002) 225–241 231
duce a peneplain are only one of a number of possibil-
ities (none of which can persist indefinitely; Phillips,
1995a). Further, the modes that could lead toward a
peneplain would eventually trigger isostatic adjust-
ment, which could move the landscape into another
state. This suggests that the relationship between ero-
sion (or deposition) and isostatic uplift (or subsidence)
should be accounted for in any consideration of relief
This assertion that isostatic compensation should
be incorporated is consistent with a number of other
studies. Vertical crustal movements may be related to
active tectonism or passive isostatic rebound response
to erosional unloading. Lambeck and Stephenson
(1986) argue that the former is dominant during active
orogenic phases, but later (i.e., during the phases in
which peneplanation might occur), the second is
dominant. Montgomery’s (1994) results from the
Sierra Nevadas and Himalayas are consistent with
that assessment, showing significant but minor con-
tributions of isostasy to total uplift in those active
orogens. The conceptual model of Masek et al. (1994)
(applied to eastern margin of central Andean plateau
and to the southern margin of Tibetan plateau) starts
with kinematic tectonic uplift, leading to a cycle of
increased orographic precipitation, greater erosion
deposition, and isostatic compensation.
Bishop and Brown (1992) took issue with some of
Lambeck and Stephenson’s (1986) results with respect
to landscape evolution in SE Australia but concurred
on the importance of isostatic rebound. In the Lachlan
River valley, a range of evidence is consistent with
isostatic response to unloading, even where denudation
rates were very low (Bishop and Brown, 1992). The
Gipfelflur hypothesis proposed by Penck (1924) and
tested successfully in the mountains of Brunei by
Dykes and Thornes (1996) also depends on an isostatic
response to denudation. About half the mean elevation
of Britain results from isostatic adjustment to denuda-
tion, with deviations from elevations expected from
unloading/rebound relationships attributable to neo-
tectonic movements and to variable rock resistance
and base levels (Clayton and Shamoon, 1999).
Isostatic responses to denudation and deposition
have a significant or even central role in most recent
models of long-term landscape evolution that attempt
to incorporate both uplift and denudation (Ahnert,
1970, 1984; Morisawa, 1975; Lambeck and Stephen-
son, 1986; Pinet and Sourian, 1988; Pitman and
Golovchenko, 1991; Bishop and Brown, 1992; Mont-
gomery, 1994; Zhou and Stu
¨ne, 1994; Gunnell, 1998).
Table 1
Possible states of relief development
(1) Higher and lower sites both eroding; higher sites at a rate
greater than or equal to lower sites. Stable. This state would
lead to a Davisian peneplain.
(2) Higher and lower sites both accreting; higher sites at a rate
less than or equal to lower sites. Stable. Applies only to
depositional landscapes.
(3) Higher and lower sites both uplifting; lower sites at a more
rapid rate. Stable. Not realistic at a regional scale.
(4) Higher sites eroding; lower sites accreting. Stable. May lead
to a Davisian peneplain if erosion exceeds accretion, so that
denudation occurs.
(5) Higher sites eroding; lower sites uplifting. Stable. Not
realistic, particularly at regional scale.
(6) No change at higher sites; lower sites accreting. Stable.
Requires external input of sediment; applies only to
depositional landscapes.
(7) No change at higher sites; lower sites uplifting. Stable. Not
realistic, particularly at regional scale.
(8) Higher sites eroding; no change at lower sites. Stable. This
state could lead to a Davisian peneplain.
(9) Initially planar surface with uniform rates of erosion,
accretion, or uplift. Stable. Not realistic.
(10) Initially planar surface with nonuniform rates of erosion,
accretion, or uplift. Unstable.
(11) Higher and lower sites both eroding; lower sites at a rate
greater than higher sites. Unstable. Incision-dominated
(12) Higher and lower sites both accreting; higher sites at a
rate greater than lower sites. Unstable. Applies only to
depositional landscapes, and not realistic.
(13) Higher and lower sites both uplifting; higher sites at
a more rapid rate. Unstable.
(14) Higher sites uplifting; lower sites eroding. Unstable. May
occur with a combination of fluvial incision and isostatic
(15) Higher sites accreting; lower sites eroding. Unstable; unlikely.
(16) No change at higher sites; lower sites eroding. Unstable.
(17) Higher sites uplifting; no change at lower sites. Unstable.
Not realistic, particularly at regional scale.
(18) Higher sites accreting; no change at lower sites. Unstable.
Not realistic.
‘‘Higher’’ and ‘‘lower’’ refer to initially higher and lower elevations
in the landscape. Changes refer to characteristic or average trends or
responses. Stable states represent convergent evolution where
initially dissimilar points converge, on average, toward more similar
elevations over time. Unstable states are divergent, where the
elevation differences of randomly selected locations become greater,
on average, over time (see Phillips, 1995a). Unstable modes cannot
result in a peneplain.
J.D. Phillips / Geomorphology 45 (2002) 225–241232
4.2. Models of erosion and isostatic compensation
Morisawa (1975) discussed the feedback between
erosion, elevation, and isostatic adjustment in the
context of models and theories of landscape evolution.
Erosion generally reduces elevation but also promotes
isostatic uplift. The extent of delay in this feedback,
according to Morisawa (1975), is the essential differ-
ence between the models of Davis (who favored long
delays) and Penck, whose framework called for the
more-or-less simultaneous operation of uplift and
denudation. Lags occur when isostatic adjustments
require an erosional unloading (or loading) threshold.
Morisawa also noted the possibility of positive feed-
back as denuded crustal areas continue to rise and
depositional basins subside because of isostatic
adjustments. Significantly, she also pointed out that
a steady-state equilibrium relationship between denu-
dation and uplift must be transitory and could occur
only when one of the two is increasing and the other
decreasing (Morisawa, 1975, p. 211).
One of the most influential models is that of Ahnert
(1970), who examined the relationships between
denudation, relief, and uplift in large mid-latitude
drainage basins. The model is based on several
fundamental relationships. The first relates denudation
rate (d;mka
) to mean relief (h, m):
where k
is a coefficient relating denudation to relief,
and k
is a parameter reflecting a constant amount of
denudation occurring because of slope- and relief-
independent processes. The equation is based on a
systematic relationship between slope gradient and
relief (i.e., relief as a surrogate for slope). The 20 basins
studied by Ahnert yielded a correlation coefficient of
r= 0.98 (with k
= 0.0001535 and k
= 0.01088). Uplift
from isostatic compensation (u)is
where qrepresents the mean densities of the crustal
material being removed and the upper mantle (t m
The q
value of 3.3 used by Ahnert is consistent in the
literature. Ahnert used q
= 2.5; other studies have used
values in the range of 2.5 to 2.7. Ahnert (1970) also
incorporated stream incision and tectonic uplift in his
model, concluding that at least 18.5 million years
would be required to reduce relief to 10% of its initial
value and that steady-state relief would require the
unlikely occurrence of constant uplift for more than
20 ma. The Ahnert model was tested in the South
Indian shield by Gunnell (1998), who found that
potential denudation rates from the model agreed well
with long-term rates from fission track studies. Ahnert
(1984) later established an upper limit of about 0.8 for
the ratio of the logarithms of maximum elevation and
width of mountain ranges, interpreting this as the
morphological expression of a dynamic equilibrium
between the geophysical maximum rate of uplift and
Pitman and Golovchenko (1991) also modelled the
isostatic response to erosional unloading with the
denudation rate proportional to average regional ele-
vation (YBAR). Their model suggests that without sea
level change, it would take more than 300 ma and
probably more than 450 ma to degrade a Himalayan-
size mountain belt to a peneplain (defined as YBAR
= 10 m). For an old mountain belt (YBAR = 500 m), an
additional 200 to 390 million years are required for
degradation to a peneplain. In a system where river
valleys are graded with respect to base level, the rate of
downcutting is enhanced by the rate of uplift caused by
isostatic response to regional denudation. Transgres-
sion of river valleys cannot occur until the rate of valley
uplift caused by isostatic response is equal to the rate of
sea level rise, which under typical nonglacial condi-
tions would not occur until degradation has reached
YBAR =60 m. Pitman and Golovchenko (1991) argue
that that under transgressive conditions, the decrease in
the rate of downcutting in river valleys greatly increases
the rate of degradation. As a consequence, older moun-
tain belts (YBAR V500 m) may be degraded to a
peneplain during sea level rise. Under falling sea level,
the model shows landscapes evolving toward a steady-
state surface where the degradation rate equals the rate
of sea level fall. With base (sea) level rising, the rate of
downcutting is decreased by rate of rise. The model is
applied to Davis’ venerable field problem, the Triassic
evolution of the Appalachian ridge-and-valley. Accord-
ing to the model, large-scale peneplanation can occur
only during episodes of significant sea level rise over
long periods ( > 250 m over >50 million years). Thus,
they argued, peneplanation is indicative of sea level
changes of this magnitude (Pitman and Golovchenko,
J.D. Phillips / Geomorphology 45 (2002) 225–241 233
Pinet and Sourian (1988), assuming that denuda-
tion rates are controlled by relief, estimated the time
scales to reduce initial relief to 10% or 37% to range
from 2 to 300 ma. Two preferential values, about 2.5
ma for active orogenies and about 25 ma (and more
variable) for dead orogenies, emerge from their
model. Thus, denudation rates ultimately depend on
While most studies of the interaction of exogenic
processes and isostatic response have focussed on
erosion and uplift, earth scientists generally recognize
that the process also works in reverse, i.e., depositio-
nal loading results in subsidence. Sunamura’s (1976)
feedback model of geosyncline development, for
example, showed exponentially accelerated subsi-
dence associated with the load of sediment and crustal
downwarping in depressions.
4.3. Elevation or relief ?
Some debate has occurred over whether elevation
or relief amplitude is the appropriate parameter to
express the interaction of denudation and uplift. Fol-
lowing Ahnert (1970, 1984), Gunnell (1998) argued
that relief is a better variable than mean, maximum, or
modal elevation as a control of denudation because of
a better correlation with slope. Plateaus with high
elevations but low slopes are obviously problematic
in this regard. Pinet and Sourian (1988) also favored
the use of relief.
Tippett and Kamp (1995), on the other hand, docu-
mented strong and consistent linear associations
between uplift and the mean surface, summit, and
valley elevations in the Southern Alps, New Zealand.
Pitman and Golovchenko (1991) determined that
regional mean elevation was the most important topo-
graphic variable in examining denudation/isostasy/
base level interactions. Zhou and Stu
¨ne (1994) devel-
oped a model examining the interplay between
dynamic uplift and denudation in response to simulta-
neous surface erosion and lithospheric deformation.
They assumed that erosion is a linear function of
elevation. Other models using elevation as the key
topographic variable include Lambeck and Stephenson
(1986) and Bishop and Brown (1992). Montgomery
(1994) and Dykes and Thornes (1996) are specifically
concerned with summit elevations, but their models
and field evidence do show that elevation can be an
appropriate variable (as opposed to slope or relief) in
examining erosion uplift topography interactions.
It is possible that relief variables are superior in
some respects, in that they relate more closely to slope
gradients and, thus, to denudation processes. Con-
versely, elevation variables more accurately reflect the
topographic load and the isostatic compensation.
Given a need or desire to maintain some degree of
simplicity, this study will follow those cited above and
choose one variable or the other. Elevation is used
here because of the explicit concern with topographic
loading and the isostatic response.
5. Landscape interaction model
The relationships between endogenic forces of
uplift and subsidence, the exogenic forces of denuda-
tion and aggradation, and elevation are straightforward.
Yet, when considered simultaneously, these interac-
tions may be quite complex. Other things being equal,
net erosion or denudation reduces elevation; and net
deposition or aggradation increases it. Elevation itself
provides a feedback to erosion and deposition via the
effects on potential gravitational energy, relief, and
slope. Thus, erosion is promoted by greater elevation,
and deposition is encouraged by lower elevations.
These relationships are discussed in the context of
geomorphic systems and the interplay during landform
and landscape evolution by Ahnert (1976, 1984, 1987).
The relationship between isostatic uplift and subsi-
dence and elevation are direct and obvious, with any
feedback effects operating through denudation and
aggradation. Additional complexity is attributable to
the isostatic responses to erosional unloading and
depositional loading (Morisawa, 1975; Ahnert, 1984;
Pitman and Golovchenko, 1991). In addition, geo-
morphic processes such as erosion and deposition
may be self-limiting independently of relationships
with isostasy, for example, due to exhaustion of avail-
able sediments or weathering limitations. These rela-
tionships are shown in Fig. 1.
The system shown in Fig. 1 is, in its general form,
a partially specified system. That is, the qualitative
nature of the relationships is known—whether an
increase or decrease in one component causes an
increase, decrease, or has no direct effect on another
component. The quantitative relationships—rates of
J.D. Phillips / Geomorphology 45 (2002) 225–241234
denudation, aggradation, elevation change, and iso-
static adjustment—are rarely known with any specif-
icity and certainly vary geographically and temporally.
In the context of landscape evolution, even if these
quantities can be quantified, all that can be known is
virtual rates; that is, the amount of movement,
removal, or deposition and the (approximate) time
period with no information on timing or tempo. Thus,
the partially specified system is examined here. The
system can be translated into an interaction matrix as
shown in Table 2, where the matrix entries represent
the positive, negative, or absence of direct effects of
the components on each other. External geodynamic,
climatic, and other factors that influence the system
components are, of course, significant but have no
effect on the stability of the system shown in Fig. 1 and
Table 2.
The stability of the interaction matrix in Table 2 is
given by its Lyapunov exponents, which are equiv-
alent to the real parts of the complex eigenvalues (k).
If all Lyapunov exponents/eigenvalues are negative,
the system is stable to small perturbations. This
stability indicates that following relatively small
changes or disturbances (for example, changes in
denudation associated with climate or short pulses in
tectonic activity), the deviations are damped and the
system returns to its pre-perturbation state. This fur-
ther suggests that if the system is stable in the absence
of cataclysmic changes, the landscape will persist in
its given state; for instance, progressive degradation
Fig. 1. Diagram showing the interactions among erosion, deposition, elevation, and isostatic uplift and subsidence. Positive and negative effects
refer to whether an increase or decrease in a component produces a change in the same (positive) or opposite (negative) direction in another
component. For example, an increase in erosion produces an increase in isostatic uplift and a decrease in elevation.
Table 2
Interaction matrix for Fig. 1
Erosion Deposition Elevation Isostatic uplift Isostatic subsidence
Erosion a
Deposition 0 a
Elevation a
00 0
Isostatic uplift 0 0 a
Isostatic subsidence 0 0 a
Matrix entries represent the positive, negative, or negligible influence of the row component (left) on the column component (top).
J.D. Phillips / Geomorphology 45 (2002) 225–241 235
toward a peneplain or relief increases via river down-
cutting into an uplifting surface. While no increasing
or decreasing relief trend can continue indefinitely
(Ahnert, 1970, 1984; Pitman and Golovchenko, 1991;
Phillips, 1995a), if the system is stable, any given
trend would persist until limits such as ultimate base
level are achieved or until a fundamental change in the
boundary conditions occurs.
If any Lyapunov exponent is positive (k> 0), the
system is unstable. In an unstable system, small per-
turbations persist and grow over finite time; and the
system does not recover to its previous state. Instability
of the interaction matrix of Table 1 would indicate that
changes in any system component would be likely to
cause the system to move to a new state. The qualitative
stability analysis of partially specified geomorphic and
geophysical systems is discussed in greater detail else-
where (Slingerland, 1981; Andronova and Schlesinger,
1991; Phillips, 1992, 1999).
The characteristic equation of the matrix is
k5þ ½ða11Þþða22 Þk4þ ½ða13 a31Þ
þða32a23 Þða11 Þða22Þk3
þ½fa14a43 a31g þ fða35Þa53 a31g
þfa25a53 ða32Þg þ fða24Þa43 ða32Þg
 fða22Þða13 Þa31g  fða11Þa23 ða32Þgk2
þ ½fða24Þa43 ða32 Þða11Þg
fa14a43 a31ða22 Þg
 fða15Þa53 a31ða22 Þgk¼0ð3Þ
Denoting the coefficients for each term F
, the
conditions for stability are established by the
Routh Hurwitz criteria. These criteria hold that the
real parts of the eigenvalues are all negative if and
only if:
(i) F
< 0 for all i.
(ii) F
>0, for n=5.
Full mathematical discussions of the method are
given by Cesari (1971), Logofet (1993), and Puccia
and Levins (1985). In this case, F
and F
are negative.
and F
are conditional (as is the second criterion
above), depending on the relative strengths of specific
links or chains of links in the system. Regardless of
these, F
= 0 and violates the Routh– Hurwitz criteria.
The system is, therefore, asymptotically unstable.
Perturbations and disturbances will persist and grow,
and will change the state of the landscape system.
Because the interrelationships between denudation,
accumulation, elevation, isostatic uplift, and isostatic
subsidence are inherently unstable, the general state of
the system is vulnerable to modification by small
perturbations. Changes in climate or biotic activity,
for instance, that modify the erosion/deposition
regime will result in a new system state. Changes in
isostatic/tectonic regimes will likewise result in a new
system state, as will changes in relative elevation
associated with base level changes such as sea level
rise or fall. The landscape would thus be expected to
be unable to maintain any given state in the face of a
variety of changes.
5.1. Supporting evidence?
The landscape interaction model and the instability
theory discussed above are intrinsically difficult to
test, as is any theory of inherently unobservable
phenomena such as landscape evolution over geo-
logical time scales. In a very simplistic sense, the
model is supported by the absence of peneplains.
However, the model implies more generally that no
particular mode of landscape evolution can persist
indefinitely in the face of tectonic, climate, and base
level changes. If this is correct, then:
(i) There should be few or no examples of land-
scapes (such as peneplains) that require the operation
of a single mode of topographic evolution over time
periods longer than those at which fluctuations of
climate, sea level, and tectonic activity occur.
(ii) There should be evidence that changes in
climate, sea level, and tectonic activity result in
changes in the fundamental mode of landscape evo-
lution (for example, between different states as
defined in Table 1) rather than just fluctuations in
the rates of geomorphic processes.
(iii) Where ancient landscapes or old lands (see
Twidale, 1999b) exist, they should not show evidence
of (or their existence should not require) the continual
existence of any particular mode or state of landscape
evolution throughout their history.
In general, the weight of geomorphic and geo-
logical evidence supports the three criteria above.
While there is reason to doubt the conventional wis-
J.D. Phillips / Geomorphology 45 (2002) 225–241236
dom that nearly all the surface of Earth’s surface is no
older than late Tertiary (see Twidale 1999a, 2000),
that generalization is true over much of the planet.
Moreover, the survival of ancient landscapes and
palaeoforms in Australia is attributable to high resist-
ance of some rocks, nonuniformity of weathering and
erosion, and self-reinforcing weathering mechanisms
that allow some resistant palaeoforms to persist (Twi-
dale, 2000, pp. 547 548). In addition, while the
relative tectonic stability of some Gondwanan surfa-
ces certainly aids the preservation of old landforms,
absolute tectonic stability is not necessary (Twidale,
2000, pp. 546– 547). Therefore, the existence of
ancient surfaces does not by itself imply the long
continuation of any particular mode of landscape
Some support exists for the instability model pre-
sented here, and in any case, some criteria for empirical
verification and falsification are available. However,
rigorous testing will require more information on the
frequency and duration of modes of landscape evolu-
tion and of climate, tectonic, and sea level fluctuations.
There are other problematic issues in applying and
evaluating the model, as acknowledged in the next
6. Discussion
If erosion rates are positively (and deposition rates
negatively) related to elevation, and if isostatic
responses to erosional unloading or depositional load-
ing occur on commensurate time scales to the erosion/
elevation relationships, then the topographic system is
inherently dynamically unstable. This instability
means that minor variations in initial conditions or
minor perturbations are likely to persist and grow over
finite time. This has several implications for landscape
evolution. First, it implies that initial variations (for
example, in erosional resistance or initial topography)
will be manifested over time as topographic variations
that are large compared to the initial variations. The
effects of perturbations (for instance, volcanic erup-
tions or climate change) will not only be large relative
to the magnitude of the disturbance but will persist
over much longer time scales. Generally, this dynam-
ical instability implies that no particular mode of
topographic evolution, including progressive down-
wearing to base level, can persist in the face of
perturbations to elevation fields, erosion or deposition
rates or magnitudes, or isostatic responses.
This finding—that the relationship between ero-
sion/deposition, elevation, and isostatic response is
inherently unstable—could explain the absence of
contemporary peneplains. The Late Cenozoic history
of the Earth, with variations in climate, glacial advan-
ces and retreats, sea level variations, and neotectonic
movements, has provided just the sort of disturbances
that are unlikely to allow any particular mode of
topographic evolution to persist for long periods.
Four major competing explanations for the absence
of peneplains can be distinguished. One is that pene-
plains simply take so long to form that tectonic
constancy has rarely existed long enough to allow
for peneplanation to run its course. This explanation is
entirely consistent with the instability-based theory
proposed here in the sense that the latter would hold
that significant tectonic changes would cause the
system to move to a new state, short circuiting any
peneplanation that was underway. Instability theory,
however, does not necessarily require frequent tec-
tonic activity to prevent peneplain formation, and the
tectonic unrest theory does not require dynamical
instability. The advantage of the instability theory in
this case is its ability to explain the absence of
peneplains in regions that have experienced long
periods of tectonic quiescence.
A separate argument, advanced at least since Davis,
holds that Earth has experienced active tectonic move-
ments since the late Pliocene. While this explanation
could readily explain the absence of peneplains in
areas with active Neogene movements, it cannot
account for the absence of contemporary peneplains
in other regions. Again, this explanation is not incon-
sistent with instability theory, which would predict that
recent tectonics would disrupt peneplanation. The
latter also predicts, however, that climate and other
environmental changes could have the same effect.
A third possible explanation can be inferred from
the Pitman and Golovchenko (1991) argument that a
long-term, sustained sea level rise is required to
produce a peneplain. Their model assumes that rivers
are graded with respect to sea level, that isostatic
response is regional (e.g., the same in valley and
interfluve areas), and that the rate of downcutting is
everywhere within the drainage reduced by the rate of
J.D. Phillips / Geomorphology 45 (2002) 225–241 237
sea level rise. In the context of the instability theory,
this could be viewed as a special case whereby no
significant variations or perturbations occur in the
relationships among elevation (relative to sea level),
isostatic response, and erosional unloading.
Finally, the general absence of peneplains could be
explained in a purely probabilistic sense. In the most
general terms, peneplanation has always required a
particular set of circumstances (including a hiatus in
uplift, dominantly subaerial fluvial erosion) and there-
fore could not apply to active uplifts, depositional
plains, or regions dominated by marine planation,
etchplanation, or other processes. Even in settings
that might allow peneplanation, Gournellos (1997)
and Phillips (1995a) showed that progressive down-
wearing toward base level was only one of several
possible states of topographic development, with no
particular reason to believe it would be more probable
or long-lasting than other states. The instability theory
can be viewed as a formalized extension of this idea,
suggesting that peneplanation is only one possible
state of topographic development and that none of the
possible states would be expected to persist for long
periods in the face of changes in climate, tectonics,
biological influences, and other disturbances.
Of course, numerous other issues are associated
with the instability theory of topographic develop-
ment. These must be addressed separately but bear
acknowledgement here. Fundamentally, the theory
suggests that no particular mode of topographic devel-
opment is likely to persist for long periods of geologic
time, unless it can be demonstrated that it is accom-
panied by extraordinary consistency in boundary
conditions and external forcings. This is inherently
difficult to evaluate, as it requires independent evi-
dence of several different geological and paleoenvir-
onmental variables. Questions also arise about the
geologic significance of changes in system state. For
example, dissection of a plateau may be disrupted by
periods of valley aggradation or limited erosion. If the
cumulative magnitude or effects of the nondissection
episodes are small compared to that of the dissection,
are they significant in a landscape evolution context?
Are they more properly viewed as secular variations
in a long-term trend rather than fundamental changes
in system state?
Interpretive issues also exist with the instability
theory, related to rates and scales. What constitutes a
‘‘small perturbation’’? Earth scientists would agree
that landscapes are certainly sensitive to large pertur-
bations that fundamentally alter the environmental
constraints or external forcings of a landscape. Insta-
bility theory holds that, additionally, the landscape is
sensitive to any perturbation of any component (ero-
sion, deposition, elevation, isostasy). However, in
dealing with time scales of landscape evolution, surely
some short-term perturbations (for instance, the
effects of a hurricane or an invasion of beavers) that
have major geomorphic impacts on time scales of
years to centuries may not be significant in a geologic
context. Methods for available for determining the
extent to which processes operating over different
Table 3
Explanations for the absence of Peneplains in the context of the proposed instability theory (see text for further discusssion)
or theory
Similarities with instability theory Differences with instability theory
Tectonic quiescence Requirement of constant constraints and
forcings for long periods, i.e., tectonic quiescence
Instability theory requires constancy of
all major forcings (not just tectonics);
tectonic quiescence does not require
dynamical instability
Neogene tectonics Both predict that recent tectonics would prevent
peneplain formation
N.T. explanation cannot be applied to
tectonically stable regions
Sea level rise Both require long periods of constancy in
relationships between base level, elevation,
erosion, and isostatic response
Instability theory would preclude
peneplanation in many cases even
with sustained sea level rise
Probabilistic Both view peneplanation as only one of many
possible states of landscape development
Instability theory indicates that no
particular mode of development is
likely to persist for long periods
Instability theory
J.D. Phillips / Geomorphology 45 (2002) 225–241238
time scales may be independent of each other. They
have been applied thus far mainly to issues associated
with reconciling the geologically recent and rapid
changes associated with human agency and vegetation
to the generally slower and sometimes ancient tempo
of geomorphic change (Phillips, 1995b, 1997). The
application of these methods to the time scales of
landscape evolution requires information on the rates,
durations, and tempos of erosional unloading, isostatic
responses, and elevation change.
7. Summary and conclusions
Despite a great deal of effort expended in searching
for them, modern peneplains are absent or exceed-
ingly rare; and evidence for these features in the
geologic record is disputed. How can the absence of
peneplains be explained? This paper proposes that the
relationship between erosional unloading (or deposi-
tional loading), elevation, and isostatic responses is
inherently dynamically unstable. This instability
implies that any particular mode of landscape evolu-
tion as defined by the qualitative nature of the
relationship between those components is sensitive
to changes or perturbations of any magnitude that
would influence any component. Thus, variations in
tectonic forcings, climate, biotic influences, and base
level would mean that peneplanation (or any other
mode of development) would not persist.
Other explanations for the paucity of peneplains
fall into four general categories as described above
and summarized in Table 3: (1) peneplanation requires
long periods of tectonic stillstand which are rare in
earth history; (2) Earth has been characterized by
recent tectonic activity; (3) long periods of sustained
sea level rise are required for peneplanation; and (4)
peneplanation is only one of many possible modes of
landscape evolution, and there is thus a low proba-
bility of finding any given area in that particular state.
The instability theory is generally consistent with the
first three in that it requires constancy in tectonic and
base level forcings to produce a peneplain. It differs
from those explanations in its requirements of relative
constancy in all forcings. The instability and proba-
bility theories are in agreement that peneplanation is
only one of many possible states or modes of land-
scape evolution. The additional implication of the
former is that no particular mode can persist in the
face of disturbances that influence elevation fields,
erosion/deposition, or isostatic response.
C.R. Twidale made many helpful and thought-
provoking comments on the manuscript, which made
it a better paper but does not necessarily imply his
Ahnert, F., 1970. Functional relationships between denudation, re-
lief, and uplift in large mid-latitude drainage basins. American
Journal of Science 268, 243 – 263.
Ahnert, F., 1976. Brief description of a comprehensive three-dimen-
sional model of landform development. Zeitschrift fu
¨r Geomor-
phologie, Supplementband 35, 1– 10.
Ahnert, F., 1984. Local relief and the height limits of mountain
ranges. American Journal of Science 284, 1035 – 1055.
Ahnert, F., 1987. Process-response models of denudation at differ-
ent spatial scales. Catena, Supplement 10, 31 – 50.
Andronova, N.G., Schlesinger, M.E., 1991. The application of
cause-and-effect analysis to mathematical models of geophysi-
cal phenomena: 1. Formulation and sensitivity analysis. Journal
of Geophysical Research 96D, 941 – 946.
Beckinsale, R.P., Chorley, R.J., 1991. The History of the Study of
Landforms or The Development of Geomorphology: Vol. 3.
Historical and Regional Geomorphology 1890 1950. Rout-
ledge, London, 496 pp.
Bishop, P., Brown, R., 1992. Denudational isostatic rebound of
intraplate highlands: the Lachlan River valley, Australia. Earth
Surface Processes and Landforms 17, 345 – 360.
Blume, H., 1992. Color Atlas of the Surface Forms of the Earth.
Harvard University Press, Cambridge, MA, 140 pp.
Cesari, L., 1971. Asymptotic Behavior and Stability Problems in
Ordinary Differential Equations. Springer, New York 271 pp.
Chorley, R.J., Beckinsale, R.P., Dunn, A.J., 1973. The History of
the Study of Landforms or the Development of Geomorphology:
Vol. 2. The Life and Work of William Morris Davis. Methuen,
London, 874 pp.
Clayton, K., Shamoon, N., 1999. A new approach to the relief of
Great Britain: III. Derivation of the contribution of neotectonic
movements and exceptional regional denudation to the present
relief. Geomorphology 27, 173– 189.
Coltori, M., Ollier, C.D., 2000. Geomorphic and tectonic evolution
of the Ecuadoran Andes. Geomorphology 32, 1 – 19.
Cui, Z.J., Li, D.W., Wu, Y.Q., Liu, G.G., 1999. Comment on pla-
nation surface. Chinese Science Bulletin 44, 2017 – 2022.
Davis, W.M., 1898. The Triassic Formation of Connecticut. U.S.
Geological Survey 18th Annual Report 1896 1897, Pt. 2.
Washington, pp. 1 – 192.
J.D. Phillips / Geomorphology 45 (2002) 225–241 239
Davis, W.M., 1899. The peneplain. American Geologist 23, 207 –
Davis, W.M., 1905. The geographical cycle in an arid climate.
Journal of Geology 13, 381 – 407.
Davis, W.M., 1912. Die Erkla¨rende Beschreibung der Landformen
(translated by A. Ru
¨hl). Teubner, Leipzig, 565 pp.
Davis, W.M., 1922. Peneplains and the geographical cycle. Geo-
logical Society of America Bulletin 33, 587 – 598.
Davis, W.M., 1932. Piedmont benchlands and Prima¨rrumpfe. Geo-
logical Society of America Bulletin 43, 399 – 440.
Dykes, A.P., Thornes, J.B., 1996. Tectonics and relief in tropical
forested mountains: the Gipfelflur hypothesis revisited. In: An-
derson, M.G., Brooks, S.M. (Eds.), Advances in Hillslope Pro-
cesses vol. 2. Wiley, Chichester, UK, pp. 975 – 994.
Eder, N., Neubauer, F., 2000. On the edge of the extruding wedge:
neogene kinematics and geomorphology along the southern Nie-
dere Tauern, eastern Alps. Eclogae Geologicae Helvetiae 93,
81 – 92.
Gournellos, T., 1997. A theoretical Markov chain model of the long
term land form evolution. Zeitschrift fu
¨r Geomorphologie 41,
519 – 529.
Gregory, K.M., Chase, C.G., 1994. Tectonic and climatic signifi-
cance of a late Eocene low-relief, high-level geomorphic sur-
face, Colorado. Journal of Geophysical Research 99B, 20141 –
Gunnell, Y., 1998. Present, past, and potential denudation rates: is
there a link? Tentative evidence from fission-track data, river
sediment loads, and terrain analysis in the South Indian Shield.
Geomorphology 25, 135 – 153.
Hack, J.T., 1960. Interpretation of erosional topography in humid
temperate regions. American Journal of Science 285, 80 – 97.
Harbor, D.J., 1997. Landscape evolution at the margin of the basin
and range. Geology 25, 1111– 1114.
Higgins, C.G., 1975. Theories of landform development: a perspec-
tive. In: Melhorn, W.N., Flemal, R.C. (Eds.), Theories of Land-
form Development. Publications in Geomorphology, Bing-
hamton, NY, pp. 1 – 28.
Johansson, M., 1999. Analysis of digital elevation data for palae-
osurfaces in south-western Sweden. Geomorphology 26, 279 –
Johansson, M., Olvmo, M., Soderstrom, M., 1999. Application of
digital elevation and geological data in studies of morphotec-
tonics and relief—a case study of the sub-Cambrian peneplain in
south-western Sweden. Zeitschrift fu
¨r Geomorphologie 43,
505 – 520.
Kennedy, W.Q., 1962. Some theoretical factors in geomorphological
analysis. Geological Magazine 94, 304 312.
King, L.C., 1953. Canons of landscape evolution. Geological Soci-
ety of America Bulletin 64, 721 – 751.
King, L.C., 1967. Morphology of the Earth, 2nd edn. Hafner Pub-
lishing, New York, 726 pp.
Kocurek, G., Lancaster, N., 1999. Aeolian system sediment state:
theory and Mojave Desert Kelso Dune Field example. Sedimen-
tology 46, 505 – 515.
Lambeck, K., Stephenson, R., 1986. The post-Paleozoic uplift his-
tory of south-eastern Australia. Australian Journal of Earth Sci-
ences 33, 253 270.
Lidmar-Bergstrom, K., 1995. Relief and saprolites through time on
the Baltic shield. Geomorphology 12, 45 – 61.
Lidmar-Bergstrom, K., 1996. Long term morphotectonic evolution
in Sweden. Geomorphology 16, 33 – 59.
Lidmar-Bergstrom, K., 1999. Uplift histories revealed by landforms
of the Scandinavian domes. In: Smith, B.J., Whalley, W.B.,
Warked, P.A. (Eds.), Uplift, Erosion, and Stability: Perspectives
on Long-Term Landscape Development Geological Society of
London, Special Publication vol. 162, pp. 85 – 91.
Logofet, D.O., 1993. Matrices and Graphs: Stability Problems in
Mathematical Ecology. CRC Press, Boca Raton, FL, 308 pp.
Markley, M., Norris, R.J., 1999. Structure and neotectonics of the
Blackstone Hill antiform, Central Otago, New Zealand. New
Zealand Journal of Geology and Geophysics 42, 205 – 218.
Masek, J.G., Isacks, B.L., Gubbels, T.L., Fielding, E.J., 1994. Ero-
sion and tectonics at the margins of continental plateaus. Journal
of Geophysical Research 99B, 13941 – 13956.
Montgomery, D.R., 1994. Valley incision and the uplift of mountain
peaks. Journal of Geophysical Research 99B, 13913 – 13921.
Morisawa, M., 1975. Tectonics and geomorphic models. In: Mel-
horn, W.N., Flemal, R.C. (Eds.), Theories of Landform Devel-
opment. Publications in Geomorphology, Binghamton, NY, pp.
199 – 216.
Ollier, C.D., 1995. Tectonics and landscape evolution in southeast
Australia. Geomorphology 12, 37– 44.
Ollier, C.D., Pain, C.F., 2000. The Origin of Mountains. Routledge,
Palmquist, R.C., 1975. The compatibility of structure, lithology, and
geomorphic models. In: Melhorn, W.N., Flemal, R.C. (Eds.),
Theories of Landform Development. Publications in Geomor-
phology, Binghamton, NY, pp. 145 167.
Paton, T.R., Humphries, G.S., Mitchell, P.B., 1995. Soils. A New
Global View. Yale University Press, New Haven, CT, 213 pp.
Pavich, M.J., 1989. Regolith residence time and the concept of
surface age of the piedmont ‘‘peneplain’’. Geomorphology 2,
181 – 196.
Penck, W., 1924. Morphological Analysis of Landforms (English
translation, 1953, Czech, H., and Boswell, K.C., MacMillan,
Phillips, J.D., 1992. Qualitative chaos in geomorphic systems, with
an example from wetland response to sea level rise. Journal of
Geology 100, 365 – 374.
Phillips, J.D., 1995a. Nonlinear dynamics and the evolution of re-
lief. Geomorphology 14, 57– 64.
Phillips, J.D., 1995b. Biogeomorphology and landscape evolution:
the problem of scale. Geomorphology 13, 337 – 347.
Phillips, J.D., 1997. Humans as geological agents and the question
of scale. American Journal of Science 298, 98 115.
Phillips, J.D., 1999. Earth Surface Systems. Complexity, Order, and
Scale. Blackwell, Oxford.
Pinet, P., Sourian, M., 1988. Continental erosion and large-scale
relief. Tectonics 7, 563 – 582.
Pitman, W.C., Golovchenko, X., 1991. The effect of sea-level
changes on the morphology of mountain belts. Journal of Geo-
physical Research 96B, 6879– 6891.
Puccia, C.J., Levins, R., 1985. Qualitative Modeling of Complex
Systems. Harvard University Press, Cambridge, MA.
J.D. Phillips / Geomorphology 45 (2002) 225–241240
Rhoads, B.L., Thorn, C.E., 1996. Observation in geomorphology.
In: Rhoads, B.L., Thorn, C.E. (Eds.), The Scientific Nature of
Geomorphology: Proceedings of the 27th Binghamton Sympo-
sium in Geomorphology. Wiley, New York, pp. 21– 56.
Riis, F., 1996. Quantification of Cenozoic vertical movements of
Scandinavia by correlation of morphological surfaces with off-
shore data. Global and Planetary Change 12, 331 – 357.
Ritter, D.F., 1988. Landscape analysis and the search for geomor-
phic unity. Geological Society of America Bulletin 100, 160–
Slingerland, R., 1981. Qualitative stability analysis of geologic sys-
tems an example from river hydraulic geometry. Geology 9,
491 – 493.
Sparks, B.W., 1960. Geomorphology. Wiley, New York, 371 pp.
Stirling, M.W., 1990. The Old Man Range and Garvie Mountains—
tectonic geomorphology of the central Otago peneplain, New
Zealand. New Zealand Journal of Geology and Geophysics 33,
233 – 243.
Stirling, M.W., 1991. Peneplain modification in an alpine environ-
ment of central Otago, New Zealand. New Zealand Journal of
Geology and Geophysics 34, 195 – 201.
Summerfield, M.A., 1996. Understanding landscape development:
the evolving interface between geomorphology and other earth
sciences. Area 28, 211– 220.
Sunamura, T., 1976. A feedback model of geosynclinal develop-
ment. Mathematical Geology 8, 663 – 667.
Tarr, R.S., 1898. The peneplain. American Geologist 21, 351– 370.
Tikkanen, M., 1994. Suomen pinnanmuodot (The landforms of Fin-
land). Terra 106, 181– 192.
Tippett, J.M., Kamp, P.J.J., 1995. Quantitative relationships be-
tween uplift and relief parameters for the Southern Alps, New
Zealand, as determined by fission track analysis. Earth Surface
Processes and Landforms 20, 153 – 175.
Trendall, A.F., 1962. The formation of ‘apparent peneplains’ by a
process of combined laterisation and surface wash. Zeitschrift
¨r Geomorphologie 6, 183– 197.
Twidale, C.R., 1994. Gondwanan (late Jurassic and Cretaceous)
palaeosurfaces of the Australian craton. Palaeogeography, Palae-
oclimatology, Palaeoecology 112, 157 – 186.
Twidale, C.R., 1999a. Landforms ancient and recent: the paradox.
Geografiska Annaler 81A, 431– 441.
Twidale, C.R., 1999b. Oldlands: characteristics and implications
based on the Australian experience. Physical Geography 20,
273 – 304.
Twidale, C.R., 2000. Early Mesozoic (?Triassic) landscapes in Aus-
tralia: evidence, argument, and implications. Journal of Geology
108, 537 – 552.
Twidale, C.R., Bourne, J.A., 1998. The use of duricrusts and topo-
graphic relationships in geomorphological correlation: conclu-
sions based in Australian experience. Catena 33, 105 – 122.
Twidale, C.R., Campbell, E.M., 1995. Pre-Quaternary landforms in
the low latitude context: the example of Australia. Geomorphol-
ogy 12, 17 – 35.
Zhou, S., Stu
¨ne, K., 1994. Modeling of dynamic uplift, denudation
rates, and thermomechanical consequences of erosion in isostati-
cally compensated mountain belts. Journal of Geophysical Re-
search 99B, 13923 – 13959.
J.D. Phillips / Geomorphology 45 (2002) 225–241 241
... Furthermore, under certain (admittedly uncommon) circumstances, the Davisian sequence is a logical outcome (Gournellos, 1997). And the interpretation of erosion surfaces (peneplains) remains an important component of historical geology and geomorphology (Phillips, 2002a). ...
... This is exemplified by the 19th and 20th century debates on the increase or decrease of relief during topographic evolution (Goudie, 2011;Thorn, 1988), though this and similar problems have rarely been expressed explicitly in terms of convergence and divergence. Exceptions include (Gournellos, 1997;Gunnell and Louchet, 2000;Phillips, 1995Phillips, , 2002aScheidegger, 1983;Twidale, 1991). From the 1990s, studies of deterministic chaos, dynamical instability, and other forms of nonlinear complexity in geomorphic systems showed that divergent development occurs frequently in a variety of geomorphic phenomena, and at a range of spatial and temporal scales (Elverfeldt, 2012). ...
An approach to landscape and Earth surface system evolution is outlined based on the inseparability of landform, soil, and ecosystem development, versus the traditional semi-independent treatment of geomorphic, ecological, pedological, and hydrological phenomena. Key themes are the coevolution of biotic and abiotic components of the environment; selection whereby more efficient and/or durable structures, forms, and patterns are preferentially formed and preserved; and the interconnected role of laws, place factors, and history. Existing conceptual frameworks for evolution of geomorphic, soil, ecological, and hydrological systems are reviewed and contrasted with the integrated approach.
... These geomorphic features can be formed by the uplift and incision of a pre-existing low-relief landscape (preservation model) or by in situ stream capture (piracy model) (Whipple et al., 2017). In the preservation model, these ancient landscapes are often tilted and dismembered during subsequent tectonic deformational phases (Jordan et al., 1989;Phillips, 2002;Zapata et al., 2019c). ...
The construction and destruction of mountain belts exert a first-order control in ecosystems by creating bridges and barriers for populations, modifying river-drainage networks and local and regional climate patterns. Several questions including How climate and tectonics control topographic growth and decay and what is the thermal and geological record of growth and decay? remain unclear and are subject of extensive research. Here we use geological data from the Antioquia Altiplano Province (AAP) in the Northern Andes to develop a 3D thermo-kinematic model that constrains past relief and exhumation rates. Results suggest that Late Cretaceous to Paleocene collision between the Caribbean Plateau and the continental margin caused high exhumation and formed a topography higher than present-day elevations. Between the late Paleocene and Oligocene, the reduction of tectonic activity caused thermal relaxation that drove regional bedrock cooling, while climatically-driven erosion significantly reduced relief, forming low-relief surfaces. During the Miocene, deformation and limited erosion resulted in a phase of Miocene topographic growth and low exhumation that preserved and deformed the previously formed low-relief surfaces. Our results demonstrate how mountain belts grow but also decay in response to the interactions and feedbacks between climate and tectonics.
... Finally, there is also significant debate over the mechanism of planation, including whether the postulated base level is sea level or whether surfaces can form in situ at higher elevations (e.g. Phillips, 2002;Clark et al., 2006;Yang et al., 2015;Chardon et al., 2016;Guillocheau et al., 2018). It is worth emphasising that rock uplift can be driven by dynamic or tectonic uplift, but is amplified by any erosion and associated isostatic rebound (England & Molnar, 1990). ...
... A Davisian-type denudation chronology is indeed what would be expected in the case of the denudation of an uplifted landscape by dominantly fluvial processes, uninterrupted by further uplift or subsidence, and in the absence of major climate or other environmental changes. Though these conditions have likely rarely (if ever?) been fully met in Earth history over long enough periods for Davisian peneplains to form, planation surfaces do exist, some created mainly by fluvial denudation (Phillips, 2002). King (1953King ( , 1957 also developed a cyclical model quite different from Davis's, but also in the single-direction, singleoutcome category, ultimately producing a pediplain. ...
This chapter focuses on the (not necessarily final) destination of landscape evolution—the attractors that landscapes may move toward and the goal functions that govern these trajectories. Single-outcome concepts posit that landscape systems move toward a single self-perpetuating state. These include notions of progression toward climax or mature forms, stable equilibrium conditions, or self-organized critical states. Multi-outcome models include notions of alternative stable states, nonequilibrium systems, and unstable attractors. As they evolve, landscapes have plasticity defined by their degrees of freedom, and constraints imposed by limits on energy, matter, and geographical space. These can be described using concepts of a multidimensional resource or landscape evolution space. Goal functions for landscape evolution are generally based on increasing fitness, often assessed in terms of optimality hypotheses, which systems strive toward maximizing or minimizing some aspect of energy and/or mass flux. Many of these are directly or indirectly related to the least action principle and maximum entropy production. Apparent goal functions can generally be explained on the basis of emergent phenomena. Landscape systems cannot aspire to anything in a literal sense, and there exist no laws that dictate trends toward the optimal states. However, if these optimal states are associated with advantages in the formation, survival, and replication of landscape components, then trends toward the optima will frequently be observed. Emergence and general principles of selection can tie together the majority of the concepts of attractors and goal functions in landscape evolution.
... Finally, there is also significant debate over the mechanism of planation, including whether the postulated base level is sea level or whether surfaces can form in situ at higher elevations (e.g. Phillips, 2002;Clark et al., 2006;Yang et al., 2015;Chardon et al., 2016;Guillocheau et al., 2018). It is worth emphasising that rock uplift can be driven by dynamic or tectonic uplift, but is amplified by any erosion and associated isostatic rebound (England & Molnar, 1990). ...
Earth's mantle undergoes convection on million-year timescales as heat is transferred from depth to the surface. Whilst this flow has long been linked to the large-scale horizontal forces that drive plate tectonics and supercontinent cycles, geologists are increasingly recognising the signature of convection through transient vertical motions in the rock record, known as "dynamic topography". A significant component of topography is supported by lithospheric isostasy, and changes in lithospheric thermal structure are sometimes included in the definition of dynamic topography. An additional component arises from active flow within the underlying convecting mantle, and this process causes dynamic topography that has lengthscales varying from 10,000 km down to 500 km and typical amplitudes of ±1 km. Transient uplift and subsidence events are often slow, but might evolve at rates as fast as 500 m/Myr over cycles as short as ~3 Myr, leading to periodic overwriting of the geological record that results in complex interpretational challenges. Despite these difficulties, a growing number of observational and computational studies have highlighted the important role of dynamic topography in fields as diverse as intraplate magmatism, sedimentary stratigraphy, landscape evolution, paleo-shorelines, oceanic circulation patterns, and ice sheet stability. This review provides a brief overview of our current understanding of the topic and explores some basic insights that can be gained from simple three-dimensional numerical simulations of mantle convection under different convective regimes. We summarise a suite of observational techniques used to estimate dynamic topography, and finish by laying out some key unanswered questions to stimulate debate and inspire future studies.
... After Penck's English translation (1953), numerous writings on Davis appeared (JUDSON, 1960;FLEMAL, 1971;DUNN, 1973;BIROT, 1974;DAVIES, 1975), followed by book chapters on the history of geomorphology (e.g., MELHORN;FLEMAL, 1975;TINKLER, 1985;THORN, 1988;CHORLEY, 1991;SUMMERFIELD, 1991;OLLIER, 1991;RHOADS;THORN, 1998;KLEIN, 1997;KENNEDY, 2006;GOUDIE, 2011), or by specialized papers (e.g., SEVON, 1983;KLEIN, 1985;MORISAWA, 1989;STODDART, 1994;PHILLIPS, 2002;OHMORI, 2003;CLAUDINO-SALES, 2005;INKPEN;COLLIER, 2007;ORME, 2007;EBERT, 2009;GREEN et al., 2013;PHILLIPS, 2015;GUNNELL, 2020). * 61 The adverb "gradually" is quite often the sign of an unconscious form of uniformitarianism, an implicit acceptance of the non-actualist principle of "uniformity of rhythm": (GOULD, 1965;1987, p. 118-126 and 174-178;GIUSTI, 2012b, box 3, p. 29-32). ...
Full-text available
Many geomorphologists today refer to Davis and his ideas without really knowing what that implies. In the second half of the 20th century, two re-evaluations of the Davisian system were carried out, which the renewed popularity of the "peneplain" concept has led us to bring back to light and discuss. RESUMO Muitos geomorfólogos hoje se referem a Davis e suas ideias sem realmente saber o que isso implica. Na segunda metade do século XX, foram realizadas duas reavaliações do sistema Davisiano, cuja renovada popularidade do conceito de "peneplanície" nos levou a trazer de volta à luz e discutir. Palavras-chave: Davis, ciclo geográfico, peneplanície, Chorley, Klein. RÉSUMÉ De nombreux géomorphologues font aujourd'hui référence à Davis et à ses idées sans vraiment savoir ce que cela implique. Dans la seconde moitié du XXe siècle, deux réévaluations du système davisien ont été effectuées, que la popularité renouvelée du concept de «pénéplaine» nous a amenées à remettre en lumière et à discuter.
... However, some scholars have questioned the term "quasi-plain" because the modern peneplains are either absent or exceedingly rare [4,5], with the main argument being whether tectonics could remain stable over a suitably extended period. A balanced uplift will never cause the area where such ...
Full-text available
Extensive areas with low-relief surfaces that are almost flat surfaces high in the mountain ranges constitute the dominant geomorphic feature of the Three Gorges area. However, their origin remains a matter of debate, and has been interpreted previously as the result of fluvial erosion after peneplain uplift. Here, a new formation mechanism for these low-relief surface landscapes has been proposed, based on the analyses of low-relief surface distribution, swath profiles, χ mapping, river capture landform characteristics, and a numerical analytical model. The results showed that the low-relief surfaces in the Three Gorges area could be divided into higher elevation and lower elevation surfaces, distributed mainly in the highlands between the Yangtze River and Qingjiang River. The analyses also showed that the rivers on both sides of the drainage divide have not yet reached equilibrium, with actively migrating drainage divides and river basins in the process of reorganizing. It was concluded that the low-relief surfaces in the Three Gorges area did not share a common uplift history, and neither were they peneplain relicts, but rather that the effect of “area-loss feedback” caused by river capture has promoted the formation of upland low-relief surface landscapes. A future work aims to present the contribution of accurate dating of low-relief surface landscapes.
... 35-30 Ma) surface uplift event has been proposed based on 561 regional mapping of geomorphic planation surfaces, river profile analysis, and the identification of 562 a regional offshore erosional unconformity at this time. Interpreted planation surfaces are typically 563 low relief to flat, weathered surfaces, however, their preservation over geological timescales has 564 been questioned (Gilchrist and Summerfield, 1991;Phillips, 2002) and their use as markers for 565 landscape evolution debated in other settings (e.g. Green et al., 2013, Pedersen et al., 2016, 566 Egholm et al., 2017. Recent work has revised the identification and classification of African 567 planation surfaces and other geomorphic features (see Guillocheau et al., 2018;Dauteuil et al., 568 2015; Picart et al., 2020). ...
Full-text available
Products of onshore passive continental margin erosion are best preserved in offshore sedimentary basins. Therefore, these basins potentially hold a recoverable record of the onshore erosion history. Here, we present apatite fission track (AFT) data for 13 samples from a borehole in the southern Walvis basin, offshore Namibia. All samples show AFT central ages older or similar to their respective stratigraphic ages, while many single grain ages are older, implying none of the samples has been totally annealed post‐deposition. Furthermore, large dispersion in single grain ages in some samples suggests multiple age components related to separate source regions. Using Bayesian mixture modelling we classify single grain ages from a given sample to particular age components to create ‘subsamples’ and then jointly invert the entire dataset to obtain a thermal history. For each sample, the post‐depositional thermal history is required to be the same for all age components, but each component (‘subsample’) has an independent pre‐depositional thermal history. With this approach we can resolve pre‐ and post‐depositional thermal events and identify changes in sediment provenance in response to the syn‐ and post‐rift tectonic evolution of Namibia and southern Africa. Apatite U‐Pb and compositional data obtained during the acquisition of LA‐ICP‐MS FT data are also presented to help track changes in provenance with time. We constrain multiple thermal events linked to the exhumation and burial history of the continental and offshore sectors of the margin over a longer timescale than has been possible using only onshore AFT thermochronological data.
Cambridge Core - Geomorphology and Physical Geography - River Dynamics - by Bruce L. Rhoads
Full-text available
The continental margin of West Greenland is similar in many respects to other elevated, passive continental margins (EPCMs) around the world. These margins are characterised by extensive regions of low relief at elevations of 1–2 kilometres above sea level sloping gently inland, with a much steeper, oceanward decline, often termed a 'Great Escarpment', terminating at a coastal plain. Recent studies, based on integration of geological, geomorphological and thermochronological evidence, have shown that the high topography of West Greenland was formed by differential uplift and dissection of an Oligo-Miocene peneplain since the late Miocene, many millions of years after continental break-up between Greenland and North America. In contrast, many studies of other EPCMs have proposed a different style of development in which the high plateaux and the steep, oceanward decline are regarded as a direct result of rifting and continental separation. Some studies assume that the elevated regions have remained high since break-up, with the high topography continuously renewed by isostasy. Others identify the elevated plains as remnants of pre-rift landscapes. Key to understanding the development of the West Greenland margin is a new approach to the study of landforms, stratigraphic landscape analysis, in which the low-relief, high-elevation plateaux at EPCMs are interpreted as uplifted peneplains: low-relief surfaces of large extent, cutting across bedrock of different age and resistance, and originally graded to sea level. Identification of different generations of peneplain (re-exposed and epigene) from regional mapping, combined with geological constraints and thermochronology, allows definition of the evolution leading to the formation of the modern-day topography. This approach is founded particularly on results from the South Swedish Dome, which document former sea levels as base levels for the formation of peneplains. These results support the view that peneplains grade towards base level, and that in the absence of other options (e.g. widespread resistant lithologies), the most likely base level is sea level. This is particularly so at continental margins due to their proximity to the adjacent ocean. Studies in which EPCMs are interpreted as related to rifting or break-up commonly favour histories involving continuous denudation of margins following rifting, and interpretation of thermochronology data in terms of monotonic cooling histories. However, in several regions, including southern Africa, south-east Australia and eastern Brazil, geological constraints demonstrate that such scenarios are inappropriate, and an episodic development involving post-breakup subsidence and burial followed later by uplift and denudation is more realistic. Such development is also indicated by the presence in sedimentary basins adjacent to many EPCMs of major erosional unconformities within the post-breakup sedimentary section which correlate with onshore denudation episodes. The nature of the processes responsible is not yet understood, but it seems likely that plate-scale forces are required in order to explain the regional extent of the effects involved. New geodynamic models are required to explain the episodic development of EPCMs, accommodating post-breakup subsidence and burial as well as subsequent uplift and denudation, long after break-up which created the characteristic, modern-day EPCM landscapes.
Full-text available
This book provides a theory to overcome the problem of identifying the principles behind the interdependence of different aspects of nature. Climate, vegetation, geology, landforms, soils, hydrology, and other environmental factors are all linked. Many scientists agree that there must be some general principles about the way in which earth surface systems operate, and about the ways in which the interactions of the biosphere, lithosphere, hydrosphere, and atmosphere manifest themselves. Yet there may be inherent limits on our ability to understand and isolate these interactions using traditional reductionist science. The argument of this book is that the simultaneous presence of order and chaos reflects fundamental, common properties of earth surface processes and systems. It shows how and why this is the case, with examples ranging from evolutionary and geological times scales to microscale examinations of process mechanics.
Full-text available
The Neogene structural and geomorphological evolution along the northwestern edge of an eastward extruding wedge within central Eastern Alps has been studied. There, the Karpatian Tamsweg basin formed at the releasing overstep between an ESE-directed normal fault exposing Lower Austroalpine units in the footwall, and an E-trending sinistral oblique-slip Prebersee fault. This fault separates deeper Middle Austroalpine units exposed in the Niedere Tauern from structurally higher Middle Austroalpine and Upper Austroalpine units (Gurktal nappe complex) adjacent to the south. Fault patterns along the Oberwolz-Tamsweg wrench corridor reflect complicate kinematics, from initial strike-slip displacement, subsequent N-S extension to late-stage, post-Karpatian faults and basin inversion due to transpression according to overall contraction within the Eastern Alps. The geomorphological evolution, including the Cenozoic formation of peneplain surfaces preserved within the extensional wedge, reflects the Neogene extrusion and surface subsidence of the extrusional wedge. This strongly contrasts with uplift of the Niedere Tauern footwall block that is characterized by a steep, immature relief.
In this paper probabilistic methods are applied to study the long term landform evolution. Especially here Markov chains are used. Markov chains can predict the states of a system in the future based on its past history. These mathematical structures are flexible stochastic methods for modelling natural processes. The state of the chains are numerous descriptive landform stages with certain geometrical characteristics. These stages follow each other in an evolutionary probabilistic sequence. This means that from an initial relatively flat relief and taking into consideration erosion, the magnitude of the uplift and the subsidence and time, many possible landform differentiations can develop. When the whole relief is eroded, the absorbing state is reached and the degradational process stops. We proposed two simple applications of this model from two contrasting geotectonic situations, which reflect different successions of the erosional stages.
Precambrian crystalline bedrock determines the major features of the present relief of Finland while the influence of overlying glacial and postglacial deposits is much less marked. Weathering and erosion processes have worn down originally mountainous landscape to a low-lying peneplain. About 80% of Finland may be classified as lowland, lying below 200 metres, and the highest point in the country is Halti, at 1328 m a.s.l. Structurally controlled valleys, fault-line scarps, quartzite monadnocks, rounded horsts and roches moutonnees are typical bedrock forms in the country. Six circular remnants of old meteorite impact craters have also been found in the ancient bedrock. The eroded bedrock surface is usually covered by till deposits, most commonly 3-4 m in thickness, and there are numerous large drumlin and moraine fields in different part of the country. Glaciofluvial eskers, deltas and kame fields are also basic elements of the relief, but the large end formations, such as the Salpausselkas and Central Finland ice marginal formation, are the most famous glacial accumulation forms in Finland. After deglaciation, clay plains with river valleys, beach and dune formations and peat bogs have developed during different phases of the postglacial era. The human role in creating landforms and modifying the operation of geomorphological processes is included as a matter of increasing importance. -from English summary
The landform development model program SLOP3D, which is based on the mass balance concept, is used to investigate 1) the interaction between weathering and denudation at a point on a slope, 2) the relationships between slope form, mean denudation rate and summit denudation rate and 3) the height limit of young mountain ranges as a function of their width. -from Author
The past decade has revealed a growing awareness amongst geomorphologists of the importance of tectonics to an understanding of long-term landform development. Similarly, there has been an increasing appreciation by geophysicists and geologists of the role played by surface geomorphic processes in such fundamental phenomena as mountain building and the landscape response to supercontinent fragmentation.