Vertical motions of the fjord regions of central East Greenland:
Impact of glacial erosion, deposition, and isostasy
Sergei Medvedev Physics of Geological Processes, University of Oslo, 0316 Oslo, Norway
Ebbe H. Hartz Physics of Geological Processes, University of Oslo, 0316 Oslo, Norway, and Aker Exploration,
Haakon VII’gt 9, 4005 Stavanger, Norway
Physics of Geological Processes, University of Oslo, 0316 Oslo, Norway
Yuri Y. Podladchikov
ABSTRACT
The origin of the vertical motions of East Greenland is a longstanding enigma. The area is characterized by a mountain chain
>1000 km long, wherein the high peaks (2.5–3.7 km elevation) reside
above a relatively thin crust (30–35 km). These mountains contain
Mesozoic marine sediments uplifted to 1.2 km during the tectonic
quietness of the middle to late Cenozoic. This uplifted area is cut
by some of the world’s biggest fjords. Scoresbysund fjord is as wide
as 60 km, and cuts 400 km into the land and more than 4 km down
from the peaks of the region. We test the potential amount of regional
uplift and subsidence caused by the fjords’ incision and associated
deposition of sediments offshore. These tests are based on modeling
the erosion process backward in time. The fjords are filled back
to the summit surface while similar weights of sediments are removed
from the shelf. The model considers the isostatic response of the lithosphere due to the loading and unloading of bedrock, sediments, water,
and ice. Our estimates show that an average of almost 1.2 km of bedrock was eroded in the region from the middle Cenozoic summit surface. Most of the erosional products were deposited on the continental
shelf outside the mouth of the fjords. Our calculations demonstrate
that rocks in the central Fjord Mountains may be uplifted as much
as 1.1 km due to the erosional unloading and flexural isostatic effects.
Thus these effects should present a main part of the mechanisms
responsible for the Cenozoic uplift in central East Greenland. Because
the North Atlantic is rimmed by young glacially carved mountain
chains, the model may be applicable to other parts of the area.
Keywords: glacial erosion, isostasy, Greenland, numerical modeling.
INTRODUCTION
The coastal landscapes of the fjords of the North Atlantic Ocean are
remarkable geological enigmas. Long chains of mountains mark the coastlines of Norway, east Canada, and Greenland (Fig. 1). The East Greenland
mountain chain has a size comparable to large orogenic systems such as
the European Alps, yet the landscape is younger than the last orogenesis
(ca. 400 Ma) and subsequent Mesozoic rift and early Cenozoic breakup
events (Henriksen et al., 2000). Although the rocks of East Greenland have
been studied extensively (e.g., Henriksen et al., 2000), there is no consensus on the origin of the mountains of East Greenland (Haller, 1971).
In central East Greenland the issue becomes particularly noticeable. Mesozoic marine sediments (Fig. 2A) are present at elevations up to
1.2 km (Haller, 1971). What lifted the marine sediments to such elevations
at times of apparent tectonic quiescence? Several mechanisms may be
responsible for such uplift. Here we consider only one scenario and show
that it can explain the major part of vertical motions in the area.
The landscape of central East Greenland is significantly shaped
by glacial carving (e.g., Odell, 1937, Fig. 2B). Several recent studies of
deeply eroded mountains around the world show that erosion-induced
isostatic rebound may be a major mechanism responsible for peak uplift
(Molnar and England, 1990; Montgomery, 1994; Pelletier, 2004; Stern
et al., 2005; Champagnac et al., 2007). The localized isostatic disequilibrium caused by glacial erosion triggers buoyancy forces that act from
Figure 1. General view of North Atlantic and adjacent Arctic region.
Study area is outlined by yellow rectangle. Greenland is shown without its ice sheet to emphasize significant topography of East Greenland mountain chain.
the deep Earth. These forces, however, are smoothed out by the flexural
rigidity of the lithosphere. Therefore, the isostatic readjustment results not
only in smaller subsidence of topography in the eroded fjord, but also
in the uplift of the surrounding noneroded area. Some numerical models
estimate erosion-driven uplift in detail (Champagnac et al., 2007; Stern
et al., 2005), but do not consider the deposition of eroded material and the
corresponding marginal subsidence. These models also do not consider
the dynamics of the process. The continuous interaction between isostatic
rebound and erosion and deposition may enhance vertical motions.
We chose a representative part of central East Greenland to test the
possible magnitude of erosion-induced isostatic rebound (Figs. 1 and 2B).
We first present several domains of this region based on geological and
morphological data. Then we present an analysis of how unloading by
erosion (in the fjords) and loading by ice (inland) and sediments (offshore)
may affect the landforms of the region.
OBSERVATIONS
The geology of central East Greenland displays remnants of the
Caledonian orogen, unconformably covered by Devonian–Carboniferous
continental deposits, Late Permian–Cretaceous dominantly marine
deposits, and thick Paleocene basalts (Haller, 1971; Henriksen et al.,
2000). For our reconstruction of the paleolandscapes we categorize the
region into six key domains (Fig. 3) based on topography and bathymetry
(GTOPO30 and ETOPO2; digital elevation models for the world developed by the U.S. Geological Survey and the National Geophysical Data
Center), ice sheet thickness (Bamber et al., 2001), Moho depth (SchmidtAursch and Jokat, 2005), offshore seismic images (Hamann et al., 2005),
and geological data (Henriksen et al., 2000).
Recent estimates of the crustal thickness of the region (Fig. 4) demonstrate that there are thick crustal roots, possibly representing the remains
of the Caledonian mountain belt (sensu stricto) under the main mountain
chain (with crustal roots 40–45 km thick; Schmidt-Aursch and Jokat,
2005). The crust thins both eastward and westward from this root.
© 2008 The Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or editing@geosociety.org.
GEOLOGY,
2008
Geology,
JulyJuly
2008;
v. 36; no. 7; p. 539–542; doi: 10.1130/G24638A.1; 7 figures.
539
A
B
Figure 3. Bedrock topography of study area (with ice removed). We
outline here several regional domains. (1) Caledonian foreland is
part of central Greenland depression covered by ice sheet 2–3 km
deep. (2) Main mountain chain is part of East Greenland mountain
chain and comprises continuous 2.5–3-km-elevation mountains.
Eastern boundary of this domain coincides with the tails of fjords.
(3) Fjord Mountains exhibit dramatic, +3 to –1.5 km, topographic
variations. (4) Uplifted marine sediments consist mainly of marine
deposits now lifted to 1.2 km elevation. (5) Continental shelf includes
shallow (0–800 m deep) offshore area covered by >3 km of Pliocene–
Pleistocene (5–0.01 Ma) sediments in that area (Hamann et al., 2005).
(6) Oceanic floor is separated by narrow strip of fast change in
bathymetry. Arrows (a and b) show place and direction of view of
photographs of Figure 2.
Figure 2. A: Late Permian, Triassic, Jurassic, and mid-Cretaceous
dominantly marine sediments, cut by sills and dikes of ca. 55 Ma
diabase. Marine sediments at top are elevated to 1.2 km above the
fjord in lower left corner. B: Glacier Korridoren in north-central Milne
Land. Surface is at ~2 km elevation. Glacier extends to sea level just
below field of view. The glacier enters the fjord just outside the picture; this fjord is ~1 km deep, resulting in total of 3 km of rock section. (See locations and direction of the views in Fig. 3.).
The main interest of this study is in the two domains, the Fjord
Mountains and uplifted marine sediments. The highest peaks of Greenland are here, on the relatively thin continental crust (Moho depth is
~30 km; Fig. 4). The uplifted marine rocks on the eastern side of these
Fjord Mountains represent a conservative estimate of 1.2 km of late
Cenozoic uplift (Fig. 2A). These adjacent areas are characterized by
large variations of topography, more than 3.5 km, indicating significant
erosion from the late Cenozoic paleosurface (Fig. 5). The overlap of
areas of deep fjords, uplifted marine sediments, and high peaks above
thin crust suggests erosion-driven uplift (Fig. 2).
MODEL
Our numerical model restores the pre-erosional situation in the
Scoresbysund region and evaluates the amplitude of vertical motions of
the isostatic response to the redistribution of material in the area. We aim
to get a first-order estimate of the possible topographic uplift due to the
bulk erosion in the region, while the detailed time evolution may be much
more complex and should be considered in a separate study.
In general, the model redistributes material by gradually filling up
local concave shapes of the upper surface of the inland part of the model
area with rocks of crustal density 2800 kg/m3, removing the corresponding mass of sediments with density of 2300 kg/m3 from the continental
shelf, and replacing these sediments by water. The surface then adjusts
540
Figure 4. General view of effective topography of Scoresbysund
region, with ice sheet replaced by crustal rock layer of same mass.
Black isolines—depth of Moho (in km). Yellow lines A–A and B–B
position profiles in Figure 5.
isostatically using a model of the flexural isostasy (Watts, 2001) with
a mantle density of 3300 kg/m3 and the lithosphere represented by an
elastic plate with effective thickness of 20 km (Young’s modulus of 1011
Pa and Poisson ratio of 0.5). In contrast to the free edge plate model
for the Antarctic margin of Stern et al. (2005), we consider a plate that
continues throughout the model. We use uniform elastic thickness for
the sake of simplicity.
We developed a finite-element–based suite called ProShell to construct the model. We first designed a regular plane mesh (500 × 300
elements for an area of 900 × 500 km; Figs. 2–4 use this planar mesh for
presenting the data) and then projected this mesh onto a spherical Earth
model. This procedure resulted in a three-dimensional shell-plate finiteelement mesh. We separated the model domain into onshore and offshore subdomains (Fig. 6A). In our simplified model, we approximated
GEOLOGY, July 2008
Figure 5. Variations of topography across two profiles from Figure 4. Thick black curves show actual effective topography (with
ice replaced by same mass of rocks) along the profiles. Gray ribbons illustrate variations of topography within 60-km-wide corridor
around corresponding profiles. Vertical dashed lines correspond to
domain boundaries (see Fig. 3 for domain names). Note that variations of topography are much higher in the areas of Fjord Mountains
and uplifted marine sediments.
this division by smoothing out the modern coastal line. This approximation is perhaps the largest deviation from the geological evolution
because shorelines may move in time.
During each iteration, we alternated between two directions (eastwest and north-south) to find and gradually fill all cross-sectional inland
concave shapes. We estimated the amount of material needed to fill the
concave shapes located east from the drainage divide. We removed
the same mass of material from the domain’s continental shelf. We then
calculated the isostatic response to this mass redistribution using the
mixed shear plate formulation (Kwon and Bang, 2000). The east and
west boundaries of the plate were fixed, whereas the other sides slip free.
The calculations were stopped when the surface was smooth and mass
redistribution insignificant (Fig. 6B).
Figure 6. General view of Scoresbysund region. A: Initial position
(recent topography of the region. B: Final result of the model with
smoothed topography and replaced offshore sediments. Yellow
line indicates drainage divide and black line separates offshore and
onshore parts of model.
RESULTS
Filling fjords and the surrounding highly eroded areas requires a significant amount of material, as much as 3 km thickness (Fig. 7A). The
average material thickness needed to smooth out inland topography on
the east from drainage divide is ~1.2 km. The isostatic response to this
redistribution of material is presented in Figure 7B. The replacement
of material resulted in significant subsidence of the inland coastal area,
accounting for as much as 1.1 km.
Most of the parameters used in the model are well established and
can vary only slightly. However, the elastic thickness of the lithosphere
cannot be estimated directly and it may affect results significantly. We
varied elastic properties of the lithosphere in the model and found that
thickness variations between 25 and 15 km result in only minor variations in the results (<100 m). Thus, the model cannot be used to estimate
effective elastic thickness in the region. Other mechanisms, such as local
deglaciation (Hansen, 2001), may be as efficient as variations of the
elastic strength of the lithosphere in the region. If the elastic thickness
is decreased down to 10 km, however, the result changes significantly,
with local subsidence of 1.5 km beneath the Scoresbysund fjord. Nevertheless, even these large variations in results do not change the main
conclusion of our study: erosion induced uplift is a major contributor to
land formation in the region.
The modeled isostatic response to erosion and deposition (Fig. 7)
does not account for the loading and unloading caused by the Greenland ice sheet. We performed additional numerical experiments to estimate the influence of the ice sheet to 3 km in thickness. We found that
although the inland load of ice may force up to 1.5 km of subsidence
along the western boundary of the model, the edge effect of loading
(bulging) is insignificant (up to 25 m) in the fjord area.
An important model assumption is that the borders of the continental shelf domain do not change their position during the experiment.
The other simplification is that the models do not tie the source to the
sink with a predefined drainage pattern. As a consequence of that, more
sediments were removed from the wide continental shelf in the north and
more rocks were put into the large fjords in the south. The consequence
of this is an underestimation of the subsidence of the northern part of
marine sediments (Fig. 3).
Two main features of our model differ from previous comparable
studies elsewhere (e.g., in Antarctica and the EuropeanAlps; Champagnac et al., 2007; Stern et al., 2005). (1) Because we place rocks
back over large areas (e.g., the Scoresbysund fjord is larger than any
fjord in Antarctica and comparable to the characteristic wavelength of
lithospheric bending), we consider dynamic interaction with isostasy.
If rocks are placed back into the fjords in a single modeling event, the
additional uneven isostatic subsidence caused by this load is ignored.
In our iterative model this effect increases subsidence in the study area
by ~10%. (2) The continental shelf adjacent to the study area is filled
with sediments, which should be taken into account in mass redistribution. Although the effect of offshore sediment removal accounts for
<5% of the changes in the subsidence of the fjord area, it does represent a more realistic geological scenario.
The low sensitivity of the onshore response to the sediment removal
reduces the importance of an accurate model for the sediment offshore
GEOLOGY, July 2008
541
The North Atlantic and Arctic regions have been characterized by
late Cenozoic vertical motions, active glacial erosion, and fjord systems
(e.g., West Greenland; Japsen et al., 2006). Therefore, the results of our
model may be applicable to the entire Arctic realm and perhaps to tectonically quiet coastal ranges globally.
ACKNOWLEDGMENTS
Reviews by J.-D. Champagnac, J. Chalmers, and editorial assistance of
T. Niemi and G. Gisler significantly improved the manuscript. This work was
supported by a Center of Excellence and Petromaks grants from the Norwegian Research Council to Physics of Geological Processes, University of Oslo.
Medvedev also acknowledges support from a Euromargins grant to J.-I. Faleide
and research grant from Aker Exploration.
Figure 7. A: Amount of moved material includes placing of more than
3 km of rocks to smooth onshore part of model and more than 2 km
of removed continental shelf. B: Isostatic response to that loading
and unloading reaches more than 1 km of regional subsidence.
placement in our study. That supports the simplifications adopted in
our model, which resulted in an unrealistic box-like shape of the preerosional ocean floor (Fig. 6B).
CONCLUSIONS
Our numerical exercises show that redistribution of material by putting the eroded material back to its original position may result in the
isostatic subsidence of coastal areas by 1.1 km. Time reversal of this may
represent a plausible model of how the landscape evolved onto today’s
form, thereby explaining uplift of the Fjord Mountains and the Mesozoic
marine sediments by ~1.1 km, as a result of unloading the area as the
fjords formed. This uplift is caused by flexural response to localized erosion and is generally not fully supported by local isostasy. The difference
between flexural and local isostasy can be estimated from the results of
our model (Fig. 7) and is comparable in magnitude to the unspecified
tectonic uplift from thermochronological and one-dimensional isostatic
modeling of Mathiesen et al. (2000).
Even though our model is based on several simplifications, the estimation of the regional uplift by erosion is robust and based on conservative
parameters. Thus we may conclude that the active erosion of the region is
the first-order feature responsible for significant uplift of the marine sediments and adjacent Fjord Mountains.
The model also estimates the mass balance between erosion and sedimentation in the region. We can conclude that the major part of eroded
material is stored on the adjacent continental shelf. If the estimations of the
age of the sediments offshore Scoresbysund by Hamann et al. (2005) are
correct, then the major part of erosion and, correspondingly, land formation
happened during the past 5 m.y. Thus, our study can contribute to discussion on the worldwide increased erosion and sedimentation rates during the
late Cenozoic (e.g., Molnar and England, 1990; Hay et al., 2002).
542
REFERENCES CITED
Bamber, J.L., Layberry, R.L., and Gogineni, S., 2001, A new ice thickness
and bed data set for the Greenland ice sheet 1. Measurement, data reduction, and errors: Journal of Geophysical Research, v. 106, no. D24,
p. 33,773–33,780, doi: 10.1029/2001JD900054.
Champagnac, J.D., Molnar, P., Anderson, R.S., Sue, C., and Delacou, B., 2007,
Quaternary erosion-induced isostatic rebound in the western Alps: Geology,
v. 35, p. 195–198, doi: 10.1130/G23053A.1.
Haller, J., 1971, Geology of the East Greenland Caledonides: London, New York,
Interscience, 413 p.
Hamann, N.E., Whittaker, R.C., and Stemmerik, L., 2005, Geological development of the Northeast Greenland Shelf, in Doré, A.G., and Vining, B., eds.,
Petroleum geology: North-west Europe and global perspectives: Proceedings of the 6th Petroleum Conference, Geological Society [London], Bath:
Geological Society [London], p. 887–902.
Hansen, L., 2001, Landscape and coast development of a lowland fjord margin
following deglaciation, East Greenland: Geografiska Annaler, ser. A, Physical Geography, v. 83A, p. 131–144.
Hay, W.W., Soeding, E., DeConto, R.M., and Wold, C.N., 2002, The late Cenozoic uplift—Climate change paradox: International Journal of Earth Sciences, v. 91, p. 746–774, doi: 10.1007/s00531–002–0263–1.
Henriksen, N., Higgins, A.K., Kalsbeek, F., and Pulvertaft, T.C.R., 2000,
Greenland from Archaean to Quaternary: Geological map of Greenland,
1:2 500 000: Geology of Greenland Survey Bulletin, v. 185, 93 p.
Japsen, P., Bonow, M.B., Green, P.F., Chalmers, J.A., and Lidmar-Bergström, K.,
2006, Elevated, passive continental margins: Long-term highs or Neogene
uplifts? New evidence from West Greenland: Earth and Planetary Science
Letters, v. 248, p. 330–339, doi: 10.1016/j.epsl.2006.05.036.
Kwon, Y.W., and Bang, H., 2000, The finite element method using MATLAB:
Boca Raton, Florida, CRC Press, 607 p.
Mathiesen, A., Bidstrup, T., and Christiansen, F.G., 2000, Denudation and uplift
history of the Jameson Land basin, East Greenland—Constrained from
maturity and apatite fission track data: Global and Planetary Changes, v. 24,
p. 275–301.
Molnar, P., and England, P., 1990, Late Cenozoic uplift of mountain ranges and
global climate change—Chicken or egg?: Nature, v. 346, p. 29–34, doi:
10.1038/346029a0.
Montgomery, D.R., 1994, Valley incision and the uplift of mountain peaks:
Journal of Geophysical Research, v. 99, no. B7, p. 13,913–13,921, doi:
10.1029/94JB00122.
Odell, N.E., 1937, The glaciers and morphology of the Franz Josef Fjord region
of north-east Greenland: Geographical Journal, v. 90, p. 111–125, doi:
10.2307/1787126.
Pelletier, J.D., 2004, Estimate of three-dimensional flexural-isostatic response to
unloading: Rock uplift due to late Cenozoic glacial erosion in the western
United States: Geology, v. 32, p. 161–164, doi: 10.1130/G20059.1.
Schmidt-Aursch, M.C., and Jokat, W., 2005, The crustal structure of central
East Greenland—II: From the Precambrian shield to the recent midoceanic ridges: Geophysical Journal International, v. 160, p. 753–760, doi:
10.1111/j.1365–246X.2005.02515.x.
Stern, T.A., Baxter, A.K., and Barrett, P.J., 2005, Isostatic rebound due to glacial
erosion within the Transantarctic Mountains: Geology, v. 33, p. 221–224,
doi: 10.1130/G21068.1.
Watts, A.B., 2001, Isostasy and flexure of the lithosphere: Cambridge, New York,
Cambridge University Press, 458 p.
Manuscript received 20 November 2007
Revised manuscript received 14 March 2008
Manuscript accepted 22 March 2008
Printed in USA
GEOLOGY, July 2008