Geomorphology 231 (2015) 229–245
Contents lists available at ScienceDirect
Geomorphology
journal homepage: www.elsevier.com/locate/geomorph
Evolution of topography of post-Devonian Scandinavia: Effects and rates
of erosion
Sergei Medvedev a,⁎, Ebbe H. Hartz a,b
a
b
Centre for Earth Evolution and Dynamics, University of Oslo, Blindern, 0316 Oslo, Norway
Det norske oljeselskap, Bryggetorget 1, Aker Brygge, 0250 Oslo, Norway
a r t i c l e
i n f o
Article history:
Received 1 June 2014
Received in revised form 29 November 2014
Accepted 5 December 2014
Available online 17 December 2014
Keywords:
AFT age
Isostasy
Glacial erosion
Numerical modeling
a b s t r a c t
The mechanisms and timing of mountain growth in Scandinavia remain enigmatic given that the region has not
been involved in active orogenesis since the Devonian and in any large-scale tectonic activities after the NE Atlantic
breakup during the early Cenozoic. In this study we combine analysis of the vertical motions of the region caused by
(dominantly) glacial erosion during the Quaternary with Apatite Fission Track (AFT) analysis of rocks from the area.
Using numerical models, we first quantify how fluvial and glacial erosion carved out the fjords and valleys to a depth
of 2 km below the paleosurfaces. This erosional episode represents a major local weight loss and results in up to a
0.8-km uplift of rocks and up to a 0.5-km rise of local topography. These estimates only marginally depend on the
effective elastic thickness of the lithosphere and thus are robust. We show then that no correlation exists between
sample altitude and published AFT data, but that a correlation does exist between AFT age and the depth below our
constructed pre-glacial summit surface. This correlation demonstrates the robustness of the numerical erosional
model, quantifies average erosion rates during Carboniferous–Cretaceous at b10 m/My, and outlines the regions
of western Scandinavia (Lofoten and Bergen areas and Møre–Trøndelag fault complex) with atypical Mesozoic–Cenozoic evolution, probably related to regional tectonic activities.
© 2014 Elsevier B.V. All rights reserved.
1. Introduction
The origin of mountains along the North Atlantic coastlines remains
enigmatic. It is a matter of continuing debate if these mountains are
topographic remains from the Caledonian mountain belt (e.g., Nielsen
et al., 2009) or if they emerged in the Cretaceous or later (e.g., Japsen
et al., 2012). Here we address parts of this issue by estimating the influence and intensity of erosion from the late Paleozoic.
The Caledonide Orogeny (450–420 Ma) was the consequence of
the continent–continent collision of Laurentia and Baltica (Cocks
and Torsvik, 2002) and represents the most recent grand-scale
mountain-building process in Scandinavia. Collapse and rifting processes began to dismember the Caledonides shortly after their formation (Andersen, 1998). The Skagerrak-centered large igneous
province 297 Ma eruption event in NW Europe covered vast areas,
estimated to at least 0.5 million km2 and thus affected the thermal
history of southern Scandinavia (Torsvik et al., 2008). The prolonged
continental rifting process between the North American–Greenland
craton and Eurasia (with several rift phases dated from late Paleozoic
until early Cenozoic; Faleide et al., 2008) led to ocean formation in
the North Atlantic at ∼ 56 Ma. The area has not been involved in
large-scale tectonic processes since.
⁎ Corresponding author. Tel.: +47 22856112; fax: +47 22855101.
E-mail address: sergei.medvedev@fys.uio.no (S. Medvedev).
http://dx.doi.org/10.1016/j.geomorph.2014.12.010
0169-555X/© 2014 Elsevier B.V. All rights reserved.
The evolution of the post-Devonian Scandinavian topography was
studied using subregional stratigraphic landscape analysis and field observations of paleosurfaces (Bonow et al., 2007; Gabrielsen et al., 2010a;
Lidmar-Bergstrom et al., 2000, 2007, 2013; Reusch, 1901; Riis, 1996).
The amplitude of vertical movements inferred from such observations,
however, is debatable (see for example discussion in Gabrielsen et al.,
2010a,b; Nielsen et al., 2010b; Olesen et al., 2013). Extending the geometry and timing of such paleolandscapes to all of Scandinavia also remains speculative. Our study method is based on numerical analysis of
modern topography and trends in thermochronological data. The method lacks the details of those above, but in contrast remains robust and
unbiased thereby permitting super-regional application.
Erosion processes can be an important mechanism to enhance relief
(Gilchrist and Summerfield, 1990, 1991; Molnar and England, 1990).
The combined effect of localized erosion and the diffused action of flexural isostasy may have resulted in a nonuniform evolution of the topography. This is the focus of recently developed models of flexural isostasy,
which study isostatic uplift resulting from surface denudation in different geological structures (Braun et al., 2013; Champagnac et al., 2007,
2009; Medvedev et al., 2008, 2013; Pelletier, 2004; Steer et al., 2012;
Stern et al., 2005). Regional studies show that the development of the
fjord system in Scandinavia strongly enhances the topography in the region (Gołędowski et al., 2013; Nielsen et al., 2009; Steer et al., 2012).
Several methods have been used to test the validity of numerical
models. Ice-related erosion partially explains the uplift of Mesozoic marine sediments along passive margins of eastern Greenland (Medvedev
230
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
et al., 2008, 2013) as well as the evolution of paleoplateaus in western
Greenland (Bonow et al., 2006; Medvedev et al., 2013). An alternative
test of models of pre-glacial topography looked at the mass balance between assumed connected onshore erosion and offshore sediments
(Steer et al., 2012). However, even considering uncertainties, analysis
showed that the Scoresbysund region in east-central Greenland lacks
offshore sediments (Medvedev et al., 2008, 2013), whereas a model of
Scandinavian erosion below the paleosurface demonstrates an excess
of offshore sediments (Dowdeswell et al., 2010; Steer et al., 2012). In
the Scandinavian realm, the area of our study, stating that the cause of
the sediment masses in the North Sea is entirely owing to the erosion
of the Norwegian mountains is challenging, as other regions (notably,
the Baltic and Central European river basins) are known to also have
contributed significant masses into North Sea sediments (e.g., Anell
et al., 2010). Here, we utilize an entirely different approach whereby
we test the erosional models for Scandinavia against published Apatite
Fission Track (AFT) data from that region.
A comprehensive database of AFT studies in Scandinavia is compiled
in Hendriks et al. (2007; continuously updated). These data were previously applied to analyze geological evolution on a subregional scale
(Hendriks et al., 2010; Redfield et al., 2004, 2005a,b). Nielsen et al.
(2009) analyzed the evolution of Scandinavian topography using a copious set of AFT data, modeling the paleolandscapes based on the thermal
history inferred from the AFT samples. The downside of such an approach is the lack of widely accepted theoretical background for interpretation of AFT data onto the thermal and exhumation history. That
type of interpretation is the subject of continuing discussions (e.g.,
Chalmers et al., 2010; Green et al., 2011; Hendriks et al., 2010; Nielsen
et al., 2010a; Redfield, 2010). In contrast, in this study, we model
paleolandscapes using the topographic data; and test this model statistically against AFT data, particularly against trends (or breaks of trends)
in AFT ages. This has the benefit that the erosion rates leading to the
modern landscape are also resolved. Our study is further augmented
by the recently published AFT data set (more than 50 samples) from
the Bergen area (Ksienzyk et al., 2014).
In this study we first present erosion backward in time as a numerical method to reconstruct pre-glacial/pre-fluvial landscapes. Applying
this method to the Scandinavian topography gives us vertical motions
caused by glacial and fluvial erosion over a younger period. Secondly,
we combine our numerical erosional model and AFT data in an analysis
to estimate early erosion rates and to test for consistency between
AFT-based erosion estimates and our hind-casted, pre-incised paleic
surfaces.
2. Numerical model of fluvial and glacial erosion
2.1. The method
Continental erosion, especially glacial carving, locally removes material from the surface and unloads the lithosphere activating the buoyancy forces from the Earth interior (Fig. 1A–B). These forces, acting on the
effectively elastic lithosphere, trigger isostatic uplift of the lithosphere.
Caused by a significant elastic strength of the lithosphere, the horizontal
extent of such uplift is usually larger than the scale of the erosion localized, e.g., within fjords; thus, isostatic readjustment results in topographic uplift of surrounding noneroded areas (Fig. 1C). Note that
removing material by erosion will always reduce the average elevation
of the subjected area, and surface uplift can occur only locally. Thus erosion cannot be a main mechanism for mountain building.
We use a simple quantitative approach by numerically filling the
eroded places with crustal material and calculating the additional
load. The resultant modeled surface is an approximation of the preerosional topography, and simplifications behind the numerical approach does not tie this process to a specific period of time. This allows
us to estimate the elastic response and potential vertical movements of
surface topography backward in time (Fig. 1D–F). The numerical model
utilizes Matlab-based numerical suite ProShell (Medvedev et al., 2008,
2013). Two grids are utilized in the model, one for the surface loads integration and another for calculation of the elastic response. The resolution of the topographic grid is 0.8 km, whereas elastic calculations were
mainly performed using a 5-km grid resolution. The topographic data is
taken from an SRTM30 digital elevation model (Becker et al., 2009).
The iterative treatment includes several numerical procedures. In
each iteration step we find concavities (simply the points within the
rectangular mesh with elevation below the average of the four neighbors) and add the material (with density of crustal rocks 2800 kg/m3)
that is required to equalize topography with the neighbors' average.
While the redistribution of material is processed on the topographic
mesh, the associated isostatic response (using a mantle density of
3300 kg/m3) is calculated on the elastic plate and is assumed to be
Fig. 1. Illustration of how erosion reshapes topography (A–C) and the model approach used in this study, erosion backward in time (D–F). We consider a simple three-layer lithosphere
(not to vertical scale) with an upper layer (gray) passively sitting on top of an elastically strong part of lithosphere (darker gray) that is in turn underlined by an inviscid asthenosphere
(light gray). The initial surface (A) is subjected to localized erosion (B) that unloads lithosphere and thus results in isostatic uplift (C). In the erosion backward in time, the localized eroded
areas, found to be concave in shape (D), are numerically filled with bedrock material (E) and then used in the calculation of the downward motion caused by the additional load (F). We
assume that absolute values of vertical motions in (C) and (F) are approximately equal. Modified after Medvedev et al. (2013).
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
immediate (Kwon and Bang, 2000; Medvedev et al., 2013). While local
cavities (within one cell of numerical mesh) may be filled within one iteration step, larger cavities require several iterations. We repeat this
process until no more impactful changes happen to the landscape.
Our modeling approach has a set of simplifications. In contrast with
previous work (Medvedev et al., 2008), we isolate the erosional effects
in this model and therefore do not strip of sediments of the same age
in the offshore regions. To reduce the number of parameters in the
model, we assume that the elastic property of the lithosphere can be described by a thin elastic plate with uniform effective elastic thickness
(EET) for the entire model domain. We compare calculations based on
231
EET ranging between 5 and 40 km, including models with nonuniform
EET (Appendix A), but use EET = 20 km in the reference model. The
value EET = 20 km may be considered a realistic representation of the
behavior of the lithosphere without significant curvature for the time
scales of several millions of years (e.g., Watts, 2001).
2.2. Results of the numerical erosional model
Here, we apply the aforementioned method to Scandinavia (Fig. 2A)
so as to estimate the pre-erosional topography (Fig. 2B). The numerical
filling of the concave topography was performed only within the
Fig. 2. Results of the application of erosion backward in time to Scandinavia. The modern topography (A) transfers into pre-erosional topography (B) by filling concavities with eroded
material (C). (D) The topography of the area subjected to uplift (positive values) and subsidence (negative) during erosion and isostatic adjustment. The black line in (A) outlines
areas subjected to erosion in the model; this area is taken to be far (at least 1000 km) away from boundaries to avoid the boundary effect. The yellow line and blue polygon in
(B) show the position of the cross section and zoomed areas from Fig. 3. The blue line in (C) indicates continent–ocean boundary (Faleide et al., 2008; Gaina et al., 2009). The results
are shown for EET = 20 km. N — Norskerenna (Norwegian channel).
232
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
boundary outlined in Fig. 2A, whereas the elastic response was calculated using the entire model domain (~500 km outside of the borders of
Fig. 2). The major fjords of southern Norway require more than
1.5 km of added material (Fig. 2C) and that additional weight results
in up to 0.7 km subsidence of the modeled plate (Figs. 3A and 4).
Transforming the backward in time modeling presented in Fig. 2A–C
to forward flow of time, we can conclude that the fjords were carved locally to a more than 1.5-km depth (from the summits to the underwater
depth of fjords), which resulted in the uplift of rocks by up to 0.7 km and
in an uplift of topography of up to 0.4 km (Fig. 2D). The erosion removes
0.4 km of rocks on average for the continental part of the region outlined
by the black line in Fig. 2A, and it can be translated into an average of
~0.2 km/My if we assume that the erosion is entirely Quaternary (last
1.5–2.5 My of the geological evolution). However, the erosion is highly
localized in the fjords and can reach up to 1 km/My (or even more)
over the same period of time (Fig. 3B). The average amount of erosion
obtained in our model corresponds well with 520 m of glacial erosion
in western Norway estimated by Dowdeswell et al. (2010).
Fig. 3A illustrates the evolution of a single profile. Two processes
control the evolution of topography in opposite ways. Whereas erosion
has the effect of deepening the Earth's surface, the isostatic adjustment
caused by the reduced weight of the crust leads to uplift of the entire
crust. Thus, the resulting topographic changes are not uniform (Figs. 3
and A.1–A.3): the surfaces of the deeply eroded fjords and valleys
move down, while the summits and plateaus move up significantly
(up to 400–500 m; Fig. 3C). Figs. 3C and A.2 show that the
Hardangervidda plateau (marked by H in Fig. 3B and C), one of the largest plateaus in Europe, has lifted up 300–400 m because of the active
erosion of surrounding fjords, which has a substantial effect on the average altitude of the 1000–1100 m plateau. This example illustrates the
significance of glacial carving on the vertical movement of the Scandinavian topography, although it does not explain the main mechanism of
mountains rise.
Appendix A shows that results of numerical model of this section
depend insignificantly on the most uncertain parameter, EET. This
demonstrates robustness of results. The following sections are
based on EET = 20 km.
Our estimations showed that glacial erosion, the major contributor
of Quaternary landscape evolution, may result locally in a significant,
up to 400-m uplift. The times of active glacial erosion also includes occurrence of more than 3-km-thick ice caps (Peltier, 2004). The vertical
motions of the lithosphere under this load can be estimated as large as
Fig. 3. (A) Evolution of topography and the elastic response for profile A–A′ (see location in Fig. 2B). Modern topography and modeled transient and pre-erosional topography displayed
using colors that correspond to the degree of erosional evolution. The elastic response is 0 at modern time and increases in absolute value with an increase of the weight of a less eroded
upper surface in the backward-in-time erosional model. Of note is the summit around 8°E, which was uplifted by more than 0.4 km. (B) The amount of erosion and (C) erosion-driven
elevation changes predicted by the model within a blue box from Fig. 2B. The results are shown for EET = 20 km. H — Hardangervidda (average elevation 1.0–1.1 km).
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
233
Fig. 4. AFT data from Hendriks et al. (2007) and Ksienzyk et al. (2014). (A) The left panel demonstrates that no correlation with the elevation of AFT samples exists, regardless of data errors.
(B) The right panel demonstrates absence of a clear pattern of AFT age depending on the geographical locations of the samples. Generally, AFT ages are younger closer to the west-Norwegian coast.
600–700 m (similar to results in Medvedev et al., 2013). We, however,
do not consider such displacements in this study as these changes
have episodic, recoverable character in contrast to permanent changes
caused by erosion.
In summary, the numerical model of this section roughly estimates
the amount of erosion and pre-erosion topography of Scandinavia filling
up concavities caused mainly by glacial and fluvial erosion. The resulting
pre-erosion topography is much smoother than the modern, and the elevation is slightly lower (note that the color scheme is the same in
Fig. 2A and B, although the extreme elevation is lower in Fig. 2B),
reaching up to 1 km in the north and locally above 1.5 km in the south
(Fig. 2B). We refer to this surface as the paleic surface, acknowledging
its development is complex, and may not have formed synchronous
all over Scandinavia.
3. Combined analysis of AFT data and erosional model
3.1. AFT data
The accuracy of the modeled pre-glacial/pre-fluvial topography is
tested using AFT data for Fennoscandia, compiled by Hendriks et al.
(2007) and recently extended by Ksienzyk et al. (2014) and presented in Fig. 4. The analysis uses more than 320 data samples collected
within the limits of our erosional model (Fig. 2A) and indicated
post-Caledonian ages (400 Ma and younger). See more details on
the data set used in the study in Appendix B.
We base our combined analysis on the amount of material added
in our erosion backward in time model (Fig. 2C); this amount is
equivalent to the amount of material eroded if we consider our
model forward in time. Owing to the limited resolution of DEM
(SRTM30, Becker et al., 2009) used in our numerical model of
Section 3, we ignore data points with recorded sample elevation
which differ significantly from the DEM model, as this mismatch
may result in significant misestimating of erosion. For each data
point our numerical model predicts a single value of exhumation;
thus, we ignore data from subvertical profiles in which data sets
with the same geographical coordinates have a variety of sample elevations and AFT ages. However, one subvertical sample profile from
Lofoten is shown in Fig. 6 and discussed in Section 3.3.2.
Attempting to find a systematic trend emerging from sampled AFT
ages, we plot them against elevation; unfortunately, no correlation
was found (Fig. 4A). The AFT ages are generally younger toward the Atlantic coast (Fig. 4B), probably relating to increased erosion along the
coast (Hendriks, 2003). The study of the Bergen area data set shows
an opposite trend, whereby samples show older AFT age coastward
(Ksienzyk et al., 2014). We, however, should mention the difference in
length-scales of observations in these two examples.
The aim of the following sections is to quantify the relation of sampled AFT ages to the depth below the enveloping (pre-erosional) surface
calculated in the numerical model of the previous section (Fig. 2B). This
depth is equivalent to the estimation of erosion (Fig. 5A) and is assigned
to each AFT sample (Fig. 5B). In the following, we term this depth as an
amount of modeled exhumation, referring to exhumation caused by fluvial and glacial incision into paleic surfaces.
3.2. Initial assumptions of combined analysis
In this section we present a set of hypothetical assumptions about
evolution of Scandinavia topography. These assumptions are then
tested using combined analysis of numerical model of Section 2
and AFT data set presented in Section 3.1 and in Appendix B.
Let us consider conditions required for AFT age data of Scandinavia
(range between 100 and 350 Ma) to be strongly correlated to the
depth below pre-erosional surface (Fig. 2B) or to the modeled
exhumation:
• Carving of the significant fjords and valleys of Scandinavia is recent
(b5–10 Ma) and thus do not change significantly the timeintegrated thermal history of samples, which is the main control
of the AFT age.
234
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
Fig. 5. (A) Results of the erosion backward in time (EBT) numerical model yields the distribution of erosion within the area (see also Fig. 2C), which can be used to estimate glacial-related
exhumation for each sample within the AFT data set (B).
• The smooth pre-erosional topography (Fig. 2B) is the dominant
shape of Scandinavia since late Paleozoic. That assumes that the
erosion was uniform until time of active glaciofluvial erosion of
the late Cenozoic. That also means that even if there were periods
of sediment accumulation, they did not accumulate a significant
amount of sediments.
• The geothermal gradient is uniform within the area. That, combined with the first two points, would result in similarity of thermal history of AFT samples located at the same depth below the
pre-erosional surface.
• No tectonic activity and thus no significant differential motions within
post-Devonian Scandinavia. This implies that Scandinavia is on the
flank of, and partly involved in, continental rifting and break-up during
the Mesozoic–Cenozoic, but that no major faulting occurs at the time
younger than AFT ages.
If the above conditions are all fulfilled, we should find a strong correlation between AFT ages and exhumation calculated from our erosion
backward in time model. That correlation in turn would give confidence
in our erosion backward in time model. The set of conditions above,
however, is highly demanding and unlikely to be fulfilled completely
over the entire Scandinavia. Deviations from any of the conditions
would result in alternation or even break of correlation. Fig. 6 demonstrates low correlation, with correlation coefficient (Bronshtein et al.,
2007) of only 0.27 and standard deviation from the trend of 0.4 km
(or 56 My). This is in contrast with our pilot study that delivered a reasonable correlation while analyzing AFT data for southern Norway only
(south from Møre–Trøndelag fault complex and without data of
Ksienzyk et al., 2014). Attempts to extend our results to all of Scandinavia, however, did not succeed (Fig. 6A).
Color coding in Fig. 6 indicates the deviation (measured in km of exhumation) of AFT data from the average exhumation-to-age line, the
general trend, for Scandinavia. The values of such differences are often
as large as the exhumation itself. The geographical distribution
(Fig. 6B) also shows low correlation between the deviation from the
general trend and the sample positions, although two observations
are noticeable:
• Three localities across Scandinavia (outlined by black boxes in Fig. 6B)
have specific properties of AFT samples in relation to the general trend
(Fig. 6A). (i) The Lofoten area (Figs. 6B and 8) is characterized by positive values of deviation from the general trend, indicating that the numerical erosional model yields exhumation that is much smaller for
the AFT age of data in this region; (ii) the area around the Møre–
Trøndelag fault complex (Figs. 6B and 9) demonstrates no correlation
to the general trend; (iii) the data around the Bergen area (Figs. 6B
and 10) are characterized by negative deviation from the general
trend demonstrating that the exhumation assigned to these data
points by the numerical model of Section 3 is too large.
• The distribution of deviation from the general trend for the remaining
data points shows a general trend of warmer colors on the south and
colder in the north (Fig. 6B).
The next section discusses these regional issues separately.
3.3. Regional scale features of the combined analysis
3.3.1. Latitudinal variations
The AFT age reflects the thermal history of the samples rather
than the exhumation process. Let us assume two samples that experience continues exhumation from two different depths, say 3 and
6 km, and with two different geothermal gradients, say 40 and 20
°C/km correspondingly. If the time taken for that exhumation is the
same, these two samples register the same AFT age, as the samples
experience the same thermal history starting from the temperature
of 120 °C, just at the border of the total annealing zone (TAZ). That
observation illustrates complications of analysis of correlation between exhumation and AFT age for extended areas with various
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
235
Fig. 6. (A) The apparent AFT age of data do not show correlation with the exhumation calculated from the numerical erosional model. Colors reflect the deviation from the general trend,
the distance in kilometers from the arbitrary line representing exhumation at a rate of 6 m/My (blue line, termed as the general trend). The black line indicates the results of the
Higravtinden profile (central part of Lofoten, see location in Fig. 8; Hendriks, 2003), where the absolute values of exhumation are not assigned within the model, but only the relative position of the data points is indicated. (B) The geographical distribution of deviation from the general trend. Black rectangles indicate the locations of the enlarged areas in the corresponding
figures. Two red polygons outline areas with surface heat flux of more than 70 mW/m2; and the blue line represents 40 mW m2 heat flux, north of which the heat flux is smaller (Slagstad
et al., 2009).
thermal conditions. Thus we introduce a corrected amount of exhumation by adding the difference in the depth of TAZ to the real exhumation. In the above example that difference is 3 km; and if we add
this difference to the first sample exhumation, the two samples
would have the same AFT age and the same corrected amount of exhumation. This allows us to compare samples from different locations; hence, if corrected exhumation is larger for one sample than
for another, we can expect that the AFT indicates younger age for
that sample. The above example is, however, extreme and results
in a significant difference between real and corrected rates of exhumation. Applied to Scandinavia, this correction is much smoother
and is realized in the form of a latitudinal correction. Moreover, the
general trend introduced in Fig. 6A allows us to avoid the use of unconstrained (paleo) geothermal gradients and absolute depths to
TOZ.
The geographical distribution of deviation from the general trend
(Fig. 6B) shows domination of negative values toward the north,
whereas the southern part of the study area demonstrates more positive deviation. The essence of this observation is presented in
Fig. 7A. The entire set of data (including small points in Fig. 7A)
does not have any preferred tendency; if we, however, take away
data points from areas around Lofoten, Trondheim, and Bergen (the
reasons for that are discussed below), the data set correlates with
linear regression dipping northward. Although it is not a perfect
match, the correlation coefficient (Bronshtein et al., 2007) of 0.5 confirms that this trend is not from random variations of data. The best
fit linear regression, ~ 40 m/° of latitude, represents the latitudinal
correction (black line in Fig. 7A).
The base latitude, 65° N, is arbitrarily assigned to zero correction.
In the adjustment procedure, the data point from, e.g., 60° N has experienced actively annealing (total or partial) conditions at a depth
200 m shallower than the average, and the new exhumation
assigned to this data point is thus 200 m larger, adjusted to the average exhumation for Scandinavia. The data point located at 70° N is
subjected to the opposite correction. Thus our analysis indicates
shallower-to-south trend and, correspondingly, higher geothermal
gradients, which can be compared with geological observations:
(i) it is compatible with higher surface heat flux measured in the
modern conditions, 60–70 mW/m 2 vs. ~ 40 mW/m 2 in the north
(Fig. 6B; Slagstad et al., 2009); (ii) it should be influenced by the
heat from the late Paleozoic Oslo plume (Torsvik et al., 2008) for at
least 50 My (e.g., Turcotte and Schubert, 2002).
Subtracting the latitudinal correction from the amount of exhumation calculated in our numerical model (Section 2) results in the
corrected amount of exhumation that we use in the following analysis.
In summary, the latitudinal correction resulted in the unbiased distribution of nonexcluded samples around the general trend (Fig. 7B).
Fig. 7A also illustrates that the three regions roughly outlined in
Fig. 7B differ significantly from the general trend for Scandinavia. This
motivates us to exclude these regions from the analysis of the general
Scandinavian trend and to consider them in separate subsections
below. Also of note is that the latitudinal correction, which decreases
the deviation from the Scandinavian average, does the opposite to
these excluded regions (Fig. 7B).
Notably, the deviation from the regional trend is measured by
vertical distance, in kilometers. However, its geological significance
may represent time difference (or, more precise, difference in apparent AFT age). Viewed as such, yellow and warmer colors represent
AFT ages that are more than 30 My younger than the regional
trend, and blue and colder colors represent samples with AFT ages
30 My or more older than the regional trend (Fig. 6 and panel B in
Figs. 8–10).
236
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
Fig. 7. (A) Distribution of the deviation from the linear law vs. latitudinal location of samples. The trend is negative in slope toward the north. The black solid line illustrates the trend and
quantifies the amount of latitudinal correction. The data points from the three special regions (small circles, outlined in Figs. 8–10) do not fit the general trend. (B) Adjusted data points and
the amount of adjustment illustrated by segments for each data point. The relation between the proximity of sample age-exhumation relation to the general trend does not depend on the
latitudinal location of the samples.
Fig. 8. Data distribution in the Lofoton–Vesterålen area. The rocks of young AFT age (mainly blue colors) in (A) are found, where our numerical model does not predict extensive erosion.
The domination of red in (B) indicates anomalous exhumation-to-age relations when compared to the general trend for Scandinavia. Modeled exhumation is corrected latitudinally.
(C) Amount of erosion predicted by the erosional model. The black star in (A) and (B) indicates the position of the Higravtinden profile (Fig. 8A; Hendriks, 2003). The blue line in
(C) indicates continent–ocean boundary (Faleide et al., 2008; Gaina et al., 2009). The black box outlines data points excluded from the analysis of the general trend for Scandinavia
(Figs. 7 and 11).
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
237
Fig. 9. Data distribution in the MFTC. Large variations of ages in the area (A) and large perturbations in the latitudinally corrected deviation from the general trend (B). (C) Amount of
erosion predicted by the erosional model. Blue lines approximately outline the major lineaments of the system: HSF — Hitra Snasa fault, VF — Verran fault; BL — Beverdalen lineament
(after Redfield et al., 2005a). The black box outlines data points excluded from the analysis of the general trend for Scandinavia (Figs. 7 and 11).
The latitudinal correction decreased the standard (square-root average) deviation from the general trend from 0.25 km or 35 My down to
0.2 km or 28 My. The analysis in the next section is based on the exhumation corrected latitudinally. Even though it is robust and supported
by independent observation, this correction is not crucial to our study,
and our main conclusions would not be altered without such a correction (see Appendix B).
3.3.2. Lofoten–Vesterålen area
The AFT ages from Lofoten–Vesterålen represent an anomaly in regard to the rest of the Scandinavian data set (Fig. 6B). With respect to
the overall trend between exhumation and age, the Lofoten–Vesterålen
data are younger and/or exhumation values assigned by the numerical
model are smaller (Fig. 8A, B). The subvertical Higravtinden profile
(Figs. 6A and 8A; Hendriks, 2003), with reasonably large mean track
length (~13.5 mm), small standard deviation of track length distribution (~1.5 mm), and similar distribution of track lengths for all the samples, indicates an approximate rate of exhumation of 30 m/My (with
even higher rates estimated for the region by Hendriks, 2003). This
could imply that the exhumation of Lofoten is younger than the rest of
Scandinavia or that exhumation is underestimated in our model. Locally, Jurassic and Cretaceous sediments can be found on-land and in
pockets between the outer Lofoten islands (Bøe et al., 2010; Dalland,
1981), suggesting that these localities were buried in the middle Mesozoic. The Lofoten islands are furthermore intensively structured by several generations of Permian to Cretaceous faults (Bergh et al., 2007;
Davids et al., 2013). Given that faulting overlaps with AFT ages, these
ages would vary across the major fault zones (Hendriks et al., 2010).
In addition, Lofoten–Vesterålen area is the region where Norwegian
land is closest to the continent–ocean boundary (Figs. 2C, 8C). The outermost continental shelf is intensely uplifted and eroded following the
55 Ma breakup of the North Atlantic (Blystad et al., 1995; Faleide
et al., 2008; Tsikalas et al., 2005). This deep Cenozoic erosion of rocks
that, in present day, sits under water west of Lofoten is not included in
our simplified numerical erosional model of Section 2, and thus the exhumation of the outermost (western) Lofoten islands is significantly
underestimated (Fig. 8C).
We suggest that the young AFT ages of Lofoten–Vesterålen samples
(Fig. 8A), the closeness of the area to the continent–ocean boundary
(Fig. 8C), and the anomalous results of application of our model to the
region (Figs. 7A and 8B) demonstrate that the evolution of Lofoten differs from the rest of Scandinavia. We can speculate that the erosion in
this area was much more intense and left only a few traces of preerosional summits. Thus, we exclude the Lofoten area from our analysis
of the general trend in Scandinavia.
3.3.3. Møre–Trøndelag fault complex (MTFC)
The Møre–Trøndelag fault complex (MTFC) area demonstrates a variety of values, in AFT age and in deviation from the general trend
(Fig. 9A and B), indicating active tectonics with multiple events. These
findings were already outlined by works of Redfield and co-workers
(Osmundsen et al., 2010; Redfield and Osmundsen, 2009; Redfield
238
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
Fig. 10. Data points within the Bergen–Sunnhordland area (mainly from Ksienzyk et al., 2014). The gray line borders the rocks related to the Bergen Arc from the outer basement rocks and
Bergdalen Nappes (Fossen, 1993). Modeled exhumation is corrected latitudinally in calculations of the deviation from the general trend (B). HSZ — Hardangerfjorden shear zone (after
Andersen et al., 1999); FF — Fenesfjorden fault (or Bergen Arc shear zone), and GF — Grimevatnet fault (after Ksienzyk et al., 2014). The black box outlines data points excluded from
the analysis of the general trend for Scandinavia (Figs. 7 and 11).
and Osmundsen, 2013; Redfield et al., 2004; Redfield et al., 2005a;
Redfield et al., 2005b).
In contrast to the Lofoten area, the AFT ages in the vicinity of the
MTFC (Fig. 9), fall above and below the general trend for Scandinavia
as a whole (Figs. 7A, 9C). Redfield et al. (2005a) illustrated how these
data fit into a local model of the late Mesozoic–Cenozoic faulting along
the MTFC. This is in agreement with the results of our study demonstrated by a large variety and amplitudes, negative and positive, of the deviation of the modeled local exhumation-age relationships from the
overall data trend in Scandinavia.
3.3.4. Bergen–Sunnhordland area
Ksienzyk et al. (2014) analyzed the morphological evolution of
the Bergen area using thermo-chronological methods and discussed
the importance of the activity of faults in the area. Our analysis supports this conclusion. Indeed, the close co-existence of the samples,
indicated by a 0 to 0.5 km deviation from the average over Scandinavia supports the presence of post-Jurassic fault activity within the region (Fig. 10A, B).
Here we can take a more general view on the AFT data analysis and
indicate that the entire area has experienced significant post-Jurassic
vertical displacement relative to the surrounding areas. Viewed in an
age perspective, this would imply that the rocks of outer BergenSunnhordland area cooled 30 to 70 My earlier than the regional trend,
further supported by occurrence of mid-Mesozoic (probably Upper Jurassic; Fossen et al., 1997) sediments within fault zones of the area.
The domination of negative deviations in Fig. 10B indicates that the
Bergen–Sunnhordland area subsided relative to inland surroundings.
Fig. 7A can give us a chance to speculate the amplitude of such a motion.
Disregarding data points along the Grimevatnet fault (several data
points with highest negative value of deviation from the general trend
in Fig. 7A), the data points within the Bergen Arc would fit into the global Scandinavian picture if the area would be uplifted by 300–500 m.
Thus, we can speculate on 300–500 m whole-region post-Jurassic
downfaulting of the Bergen–Sunnhordland area.
Although the total vertical motion may be distributed between several fault systems, two main candidates to accommodate the major regional movement seem to be the Hardangerfjorden shear zone (HSZ)
and the Fenesfjorden fault (FF) or the Bergen Arc shear zone (Fig. 10).
Although the time of FF activity is unknown, Andersen et al. (1999) reported that HSZ was active during the Cretaceous–Tertiary, well after all
the reported AFT ages from the area.
3.4. Corrected correlation between AFT ages and modeled exhumation
Analysis of the distribution of AFT data and their relation to the general trend presented in Figs. 6–10 show:
• The MTFC does not fit to our initial assumption of no tectonic activity.
The Lofoten–Vesterålen area has erosion of greater scale than one
which can be explained by our simple erosional model. The Bergen–
Sunnhordland area exhibits relative vertical motion postdating the
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
AFT ages and locally preserving Jurassic sediments. Thus, we should
exclude these areas from consideration.
• Shallower annealing conditions in the South demonstrated by our
analysis correlate with slightly higher surface heat flux measured
in present day conditions and with paleoheat from the Oslo
plume during the late Paleozoic (Torsvik et al., 2008). Thus, in the
analysis following Fig. 7B, we use an iteratively derived coefficient
of 40 m/° of latitude correction. This approximately corresponds to
the 400 m difference to a depth of active fission tracks annealing
over the entire domain, to a maximum of 4 °C/km difference in geothermal gradients.
The correlation between AFT age and modeled exhumation becomes
much stronger after the above corrections (Fig. 7). That is illustrated by
a correlation coefficient of 0.7 for this chosen set (cf. the coefficient of
0.27 for the entire set of data in Fig. 6) and standard deviation from
the general trend of 0.2 km and 28 My (cf. 0.4 km and 56 My for the
whole set in Fig. 6). The resulting correlation shows that erosion over
Scandinavia did not greatly exceed an average of 5–7 m/My, which is
characterized by the general trend.
4. Discussion
The application of the AFT data in the study of the North Atlantic
paleolandscapes is the subject of debates (e.g., Chalmers et al., 2010;
Green et al., 2011; Green et al., 2013; Japsen et al., 2013; Nielsen et al.,
2009; Nielsen et al., 2010a; Pedersen et al., 2012; Pedersen et al.,
2013; Redfield, 2010). Without entering details of the method and application of the AFT data to local geology, we demonstrated an overall
correlation between AFT data and exhumation below our modeled
paleosurface for Scandinavia. While this application clearly lacks the refinement of local studies it does represent an unbiased overall study of
post-Caledonian landscape evolution.
239
The resulting high correlation of the AFT ages and depth below
an enveloping (pre-erosional) surface (Fig. 11) indicates that
after removing three anomalous regions and adjusting for north–
south variations of geothermal gradients the four conditions listed
in Section 3.2 are satisfied. The satisfaction is, of course, approximate, which allows us to present the following discussion as
reasonable speculations rather than fully proven detailed statements. The correlation supports the assumption that the preerosional topography presented in Fig. 2B is a dominant shape of
the upper surface for Scandinavia during the late Paleozoic and Mesozoic. That includes elevation differences across Scandinavia,
which is in the order of 1 km. The high correlation between age
and exhumation also supports the assumption that the spatial distribution of erosion during the Mesozoic and late Paleozoic was
close to uniform.
In our study we do not consider individual histories of AFT samples, we did not assume absolute values of (paleo)geothermal gradients, the subjects of great uncertainties. We consider trends of AFT
ages using a new approach of combined analysis with a numerical
erosional model. The major assumption regarding the AFT data interpretation we use is the dominance of a monotonic cooling of the AFT
samples. How reasonable is this assumption? Chalmers et al. (2010)
and Japsen et al. (2013) demonstrated that similar ages and track
distributions in individual samples can be explained equally well
by monotonic cooling or by episodic heating and cooling. Application
of that observation to a large data set, however, is not trivial. Episodic
changes of significant amplitude in thermal conditions applied to a
set of samples of different initial depths would result in a situation
of some samples influenced by partial or total annealing conditions
(Green et al., 2013). That in turn would result in an increase of standard deviation (STD) of a track length distribution or even a reset of
the sample age (Green et al., 2002). The data set used in our study
has more than 200 points with a reasonably small average STD of
1.5 μm; only 7% of the data points have STD N 2 μm (Appendix B,
Fig. 11. (A) The relation between AFT ages and latitudinally corrected modeled exhumation becomes much stronger after removing areas of proposed late tectonic activity and introducing
gradient in closure depth. (B) Geographical distribution of the deviation from the linear law with excluded data points outlined by the limiting boxes.
240
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
Fig. B.1); the correlation presented in Fig. 11A excludes a significant
reset of age in AFT samples. Thus, even if episodes of thermal fluctuations were applied to the samples considered here, the amplitude of
these variations should be within the accuracy of the estimations
presented in the resulting Fig. 11A.
The accuracy of the regression, measured by the square-rooted
average of deviation from the general trend (0.2 km or 28 My),
should be added to the accuracy of data (the data used in the analyses have an error of b30 My in AFT age) and to the potential divergence caused by latitudinal correction (0.2 km). Thus, we can
conservatively assume that the paleic surface (Fig. 2B) and the general trend (Fig. 11A) are representative for late Paleozoic–Mesozoic
evolution of Scandinavian topography within an accuracy of
100 My in age and 1 km of topographic variations across the general area. That estimation is also valid while describing the monotonic cooling of the bulk of AFT samples; i.e., the episodic heating of
samples does not contradict our estimations if it involves deepening by b 1 km.
Based on these rough estimates we can speculate that the preerosional topography of Fig. 2B was varying on a kilometer scale during the late Paleozoic–Mesozoic (e.g., Gabrielsen et al., 2010a) and
that the Scandinavian topography being at sea level is possible
(e.g., Lidmar-Bergstrom et al., 2013), while mountains of 3–4 km
are unlikely (e.g., Nielsen et al., 2009). A large-scale sediment cover
is also possible, but it is unlikely to be much more than 1 km thick
over a large spatial extend because it would lead to tracks annealing
in some samples and thus AFT age resetting (Green et al., 2002;
Green et al., 2013). Note that estimation of erosion rate presented
above (5–7 m/My) reflects only the grand-scale average rate, excluding potential accumulation and erosion of sediment cover.
The above estimations consider that the grand-scale topography
evolution of Scandinavia and local violations of the predicted tendencies
do not disprove the conclusions of our study. Thus, more than 2 km of
now eroded sediment cover in the MTFC area (Weisz, 1992) is a local
phenomenon, and sediments of this thickness should not be expected
in the rest of Scandinavia according to our analysis. The effect of this
paleobasin may contribute to anomalous evolution of the MTFC region
though (see Section 3.3).
Initially, the combined analysis of the numerical erosion model
and AFT data (Section 3.2) was chiefly quantitative (Fig. 6) and did
not assume partition of the data set into subsets according to regional studies. This general approach (e.g., Fig. 7) allowed us to locate regions with an anomalous relation between AFT age and
modeled exhumation (Figs. 8–10). Our observations compare
well with observations of previous studies (e.g., Hendriks et al.,
2007, 2010; Ksienzyk et al., 2014; Redfield et al., 2005a) that indicate structural offsets within these regions. Our study was capable
of not only repeating these conclusions, but contrasting observations in these areas with data across a wider area. This approach
(e.g., Fig. 7A) helped to estimate the vertical movement of the entire Bergen–Sunnhordland area along major faults by 300–500 m,
which would be impossible studying data only within the region.
This study estimates the average erosion rate in Scandinavia of
5–7 m/My since the late Paleozoic. This estimate involves direct conversion of AFT age to the geological age, and thus the estimate is conservative, meaning that the average over time and space rate can be even
smaller.
5. Conclusions
The combined analysis of erosion model and AFT ages allows us to
conclude:
• The erosion model is robust and reasonable as a first-order regional
approximation for the reconstruction of the pre-incision paleic surface
of Scandinavia.
• The vertical motion of topography caused by fluvial and glacial
carving is a significant contributor to uplift of the Scandinavian
mountains (up to 0.4–0.5 km). This result is almost independent
from uncertain parameters such as EET.
• The average thickness of the material eroded during active fluvial
and glacial carving in Scandinavia is 0.4 km. This would correspond
to an average erosion rate of more than 150 m/My assuming the incision to be dominantly Quaternary. Locally, however, the rate may
exceed 1 km/My as a consequence of the localized nature of fluvial/
glacial erosion.
• The erosion rate during the late Paleozoic and Mesozoic (based on
available AFT ages of 325 to 100 My) is b10 m/My. This rate is low,
and thus topographic variations of Scandinavia were limited.
Our simple model allows the identification of anomalous regions
within Scandinavia (nonstandard evolution or active tectonic): Lofoten–Vesterålen, Bergen–Sunnhordland area, and the Møre–Trøndelag
fault complex.
Acknowledgments
Financial support from Det norske oljeselskap and the Norwegian
Research Council through a Centre of Excellence grant to the CEED is
appreciated. Bart Hendriks is thanked for working on and providing
the AFT data for Scandinavia. Discussions with Torgeir Andersen,
Per Terje Osmundsen, and Bart Hendriks, editorial support by Lucy
Medvedeva and, especially, by Richard A. Marston are greatly appreciated. Reviews by Tim Redfield, Peter Japsen, Paul Green, and two
anonymous reviewers helped improve the early version of the
manuscript.
Appendix A. Variations of effective elastic thickness (EET)
One of the most uncertain parameters of our numerical model of
erosion backward in time is the effective elastic thickness (EET) of
the lithosphere. The results presented in the main part are based
on constant EET = 20 km. Here we show how results depend on variations of EET. No widely accepted model for EET of Scandinavia exists, and thus we use mainly constant EET models. Some works,
however, demonstrate east–west decreasing of EET (Djomani et al.,
1999; Perez-Gussinye and Watts, 2005), although absolute values
differ significantly between models. We have analyzed models with
uniform and variable EET ranging from 5 to 40 km, considering
models involved 5 and 40 km as extreme.
Even though the flexural rigidity of the lithosphere changes up
to three orders of magnitude when elastic thickness ranging from
5 to 40 km, the results (Figs. A.1 and A.2) show only minor variations for the EET values relevant to the Earth's lithosphere in the
region (EET N 5 km). The elastic response of models with smaller
EET has larger amplitude and mimics the distinct topographic
lows such as Norskerenna and the fjords of western Norway
(Fig. A.1A). Models with higher EET result in smoother and wider
elastic response. Models with variable EET (Fig. A.1E and F) show
similarity with the model with EET = 20 km, indicating that the
mean value of variable topography controls response more than
local variations of EET.
Topography changes vary even less with EET variations (Fig. A.2).
Thus, the uplift of the Hardangervidda plateau is significant (200 m)
already for EET = 10 km, increases up to 300–400 m for EET =
15 km, and does not change for higher values of EET (Fig. A.2). The
closeness of the results for different models is even more evident
along profile A–A′ (Fig. A.3). Neither topography, nor elastic response (excluding models with extreme values of EET, 5 and
40 km) differ significantly. Fig. A.4 demonstrates that even though
the amount of exhumation may be slightly different for models
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
241
Fig. A.1. Amplitude of the elastic response to the unloading by erosion for models with uniform EET = 5, 15, 20, and 40 km (A–D) and with variable EET (E, F). The two lines, A and B,
present variations of EET: west from line A EET is 10 km in (E) and 5 km in (F); EET is 25 and 40 km correspondingly east from line B. EET is linearly matched between two border
lines. The blue color indicates a wide region of peripheral negative warping; the amplitude of this vertical motion is small as shown by the small blue negative strip on the color bars.
N — Norskerenna.
with different EETs, the correlation with AFT age stays even for
models with extreme values of EET.
This low dependence of results from the main parameter of the
model (EET) demonstrates the robustness of the estimations of the erosion backward in time model.
Appendix B. Additional analysis of AFT data
Owing to the short list of sources of AFT data, we do not present all
the data points here (Hendriks et al., 2007; Ksienzyk et al., 2014). The
total number of AFT data points that originate from the area of our
study (outlined by the black line in Fig. 2A) is 405.
We exclude 40 data points when error in estimation of apparent age
is more than 30 My or when error estimation is absent. The data are
treated together with our numerical model of erosion backward in
time, and thus accuracy of that model brings additional concern. The numerical model can only estimate a single value of exhumation for each
data point. Thus, data points of close location (within 1 km) but with
different elevations are excluded. That is mainly (sub) vertical transects
(altogether 23 data points). The 24 data points were excluded because
the DEM model used in numerical calculations differs with AFT data
by more than 350 m. Thus, we considered data of 324 points, with
219 data points having extended parameter studies (mean track length,
MTL, and standard deviation of track length distribution, STD).
Even though we base our study on the trend of the AFT apparent
ages, Fig. B.1 presents some analysis of additional characteristics of
AFT data. The data does not exhibit any correlation between these characteristics (MTL and STD) and apparent age, deviation from the general
trend, and regional location (see outlines of chosen regions in Fig. 11B).
242
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
Fig. A.2. Topography uplift during the erosion process. The fjord areas subsided significantly, while summits rose by 200–400 m. The results demonstrate that the topographic evolution
depends only a little on the (realistic, N5 km) values of EET.
Fig. A.3. Comparison of results of the erosion backward in time along profile A–A′ (Fig. 2B) for different EETs (including variable EET). Solid black line represents modern topography. Solid
lines of different colors represent pre-erosional topography for different values of EET; dash lines of the same color illustrate elastic response.
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
243
Fig. A.4. Fig. 11 of the main text is based on EET = 20 km. Variations of elastic property of the lithosphere in the model (constant EET = 5 and 40 km in (A) and (B), and variable EET with
limits 5 and 40 km in (C)) demonstrate a similar trend.
The average value of MTL is 13 μm, average STD is 1.5 μm. A significant part (88%) of samples have MTL N 12 μm and STD b 2 μm at
the same time. More than half of the samples (56%) fit into parameters defined for the unperturbed basement (Gleadow et al., 1986)
with MTL N 12.5 μm and STD b 1.6 μm.
Appendix A (Fig. A.4) demonstrates that variations of EET do not
alter the correlation presented in Fig. 11A. We discuss two additional
aspects that can affect this correlation. The latitudinal correction
(Fig. 7) may be considered as speculative; therefore we show here
that results without that correction also show correlation (Fig. B.2A).
Fig. B.1. Distribution of parameters of AFT data used in the study.
244
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
Fig. B.2. Variations of general trend and deviations from it: two special cases of data analysis. (A) The general trend for data without latitudinal correction illustrates 30% faster exhumation
than one presented in Fig. 11A. Note that correlation is not as good as in Fig. 11A. (B) The general trend for MTL-corrected age (after Cederbom et al., 2000) illustrates 24% slower exhumation. Note that the horizontal axis in this case differs from others of this type of plots. Smaller number of data points in (B) reveals the lack of MTL analysis in some data subsets.
This correlation is not as strong as in Fig. 11A and the general trend is
30% steeper, but that does not lead to negation of the main conclusions
of our study, which were taken with conservative caution.
The relation between AFT apparent age and geological ages is accepted as nontrivial and is still a matter of continuing research. One simplified approach of this relation was suggested in Cederbom et al.
(2000) based on analysis of Green (1988). In this approach the real
age of samples were derived by multiplying the AFT age by the ratio
of original track length (we use 16 μm here) to the MTL of the sample.
Presenting the results using this approach (Fig. B.2B), we do not discuss
validity or applicability of the approach of Cederbom et al. (2000), but
demonstrate that this approach does not alternate the strong correlation between exhumation of our numerical erosional model and age derived from AFT data.
References
Andersen, T.B., 1998. Extensional tectonics in the Caledonides of southern Norway, an
overview. Tectonophysics 285 (3), 333–351.
Andersen, T.B., Torsvik, T.H., Eide, E.A., Osmundsen, P.T., Faleide, J.I., 1999. Permian and
Mesozoic extensional faulting within the Caledonides of central south Norway.
J. Geol. Soc. 156, 1073–1080.
Anell, I., Thybo, H., Stratford, W., 2010. Relating Cenozoic North Sea sediments to topography in southern Norway: the interplay between tectonics and climate. Earth Planet.
Sci. Lett. 300 (1–2), 19–32.
Becker, J.J., Sandwell, D.T., Smith, W.H.F., Braud, J., Binder, B., Depner, J., Fabre, D., Factor, J.,
Ingalls, S., Kim, S.H., Ladner, R., Marks, K., Nelson, S., Pharaoh, A., Trimmer, R., Von
Rosenberg, J., Wallace, G., Weatherall, P., 2009. Global bathymetry and elevation
data at 30 arc seconds resolution: SRTM30_PLUS. Mar. Geod. 32 (4), 355–371.
Bergh, S.G., Eig, K., Klovjan, O.S., Henningsen, T., Olesen, O., Hansen, J.A., 2007. The
Lofoten–Vesteralen continental margin: a multiphase Mesozoic–Palaeogene
rifted shelf as shown by offshore-on shore brittle fault-fracture analysis. Nor.
J. Geol. 87 (1–2), 29–58.
Blystad, P., Brekke, H., Faerseth, R.B., Larsen, B.T., Skogseid, J., Torudbekken, B., 1995. Structural elements of the Norwegian continental shelf. Part II: the Norwegian Sea Region.
NPD, Norw. Petrol. Directorate Bull 8, p. 45.
Bøe, R., Fossen, H., Smelror, M., 2010. Mesozoic sediments and structures onshore Norway
and in the coastal zone. Norges Geol. Unders. Bull. 450, 15–32.
Bonow, J.M., Japsen, P., Lidmar-Bergstrom, K., Chalmers, J.A., Pedersen, A.K., 2006. Cenozoic uplift of Nuussuaq and Disko, West Greenland — elevated erosion surfaces as uplift
markers of a passive margin. Geomorphology 80 (3–4), 325–337.
Bonow, J.M., Lidmar-Bergstrom, K., Japsen, P., Chalmers, J.A., Green, P.F., 2007. Elevated erosion surfaces in central West Greenland and southern Norway: their
significance in integrated studies of passive margin development. Nor. J. Geol.
87 (1–2), 197–206.
Braun, J., Deschamps, F., Rouby, D., Dauteuil, O., 2013. Flexure of the lithosphere and the
geodynamical evolution of non-cylindrical rifted passive margins: results from a numerical model incorporating variable elastic thickness, surface processes and 3D
thermal subsidence. Tectonophysics 604, 72–82.
Bronshtein, I.N., Semendyayev, K.A., Musiol, G., Muehlig, H., 2007. Handbook of Mathematics, 3. Springer.
Cederbom, C., Larson, S.A., Tullborg, E.L., Stiberg, J.P., 2000. Fission track
thermochronology applied to Phanerozoic thermotectonic events in central
and southern Sweden. Tectonophysics 316 (1–2), 153–167.
Chalmers, J.A., Green, P., Japsen, P., Rasmussen, E.S., 2010. The Scandinavian mountains
have not persisted since the Caledonian orogeny. A comment on Nielsen et al.
(2009a). J. Geodyn. 50 (2), 94–101.
Champagnac, J.D., Molnar, P., Anderson, R.S., Sue, C., Delacou, B., 2007. Quaternary
erosion-induced isostatic rebound in the western Alps. Geology 35 (3),
195–198.
Champagnac, J.D., Schlunegger, F., Norton, K., von Blanckenburg, F., Abbuhl, L.M., Schwab, M.,
2009. Erosion-driven uplift of the modern Central Alps. Tectonophysics 474 (1–2),
236–249.
Cocks, L.R.M., Torsvik, T.H., 2002. Earth geography from 500 to 400 million years ago: a
faunal and palaeomagnetic review. J. Geol. Soc. 159, 631–644.
Dalland, A., 1981. Mesozoic Sedimentary Succession at Andoy, Northern Norway, and Relation to Structural Development of the North Atlantic Area.
Davids, C., Wemmer, K., Zwingmann, H., Kohlmann, F., Jacobs, J., Bergh, S.G., 2013. K–Ar
illite and apatite fission track constraints on brittle faulting and the evolution of the
northern Norwegian passive margin. Tectonophysics 608, 196–211.
Djomani, Y.H.P., Fairhead, J.D., Griffin, W.L., 1999. The flexural rigidity of Fennoscandia:
reflection of the tectonothermal age of the lithospheric mantle. Earth Planet. Sci.
Lett. 174 (1–2), 139–154.
Dowdeswell, J.A., Ottesen, D., Rise, L., 2010. Rates of sediment delivery from the
Fennoscandian Ice Sheet through an ice age. Geology 38 (1), 3–6.
Faleide, J.I., Tsikalas, F., Breivik, A.J., Mjelde, R., Ritzmann, O., Engen, O., Wilson, J., Eldholm,
O., 2008. Structure and evolution of the continental margin off Norway and Barents
Sea. Episodes 31 (1), 82–91.
Fossen, H., 1993. Structural evolution of the Bergsdalen Nappes, southwest Norway. Nor.
Geol. Unders. Bull. 424, 23–50.
Fossen, H., Mangerud, G., Hesthammer, J., Bugge, T., Gabrielsen, R., 1997. The Bjoroy Formation: a newly discovered occurrence of Jurassic sediments in the Bergen Arc System. Nor. Geogr. Tidsskr. 77 (4), 269–287.
Gabrielsen, R.H., Faleide, J.I., Pascal, C., Braathen, A., Nystuen, J.P., Etzelmuller, B.,
O'Donnell, S., 2010a. Latest Caledonian to Present tectonomorphological development of southern Norway. Mark. Pet. Geol. 27 (3), 709–723.
Gabrielsen, R.H., Faleide, J.I., Pascal, C., Braathen, A., Nystuen, J.P., Etzelmuller, B.,
O'Donnell, S., 2010b. Reply to discussion of Gabrielsen et al. (2010) by Nielsen et al.
S. Medvedev, E.H. Hartz / Geomorphology 231 (2015) 229–245
(this volume): Latest Caledonian to present tectonomorphological development of
southern Norway. Mark. Pet. Geol. 27 (6), 1290–1295.
Gaina, C., Gernigon, L., Ball, P., 2009. Palaeocene–Recent plate boundaries in the NE Atlantic and the formation of the Jan Mayen microcontinent. J. Geol. Soc. 166, 601–616.
Gilchrist, A.R., Summerfield, M.A., 1990. Differential denudation and flexural isostasy in
formation of rifted-margin upwarps. Nature 346 (6286), 739–742.
Gilchrist, A.R., Summerfield, M.A., 1991. Denudation, isostasy and landscape evolution.
Earth Surf. Process. 16 (6), 555–562.
Gleadow, A.J.W., Duddy, I.R., Green, P.F., Lovering, J.F., 1986. Confined fission-track lengths
in apatite — a diagnostic-tool for thermal history analysis. Contrib. Mineral. Pet. 94
(4), 405–415.
Gołędowski, B., Egholm, D.L., Nielsen, S.B., Clausen, O.R., McGregor, E.D., 2013. Cenozoic
erosion and flexural isostasy of Scandinavia. J. Geodyn. 70, 49–57.
Green, P.F., 1988. The relationship between track shortening and fission-track age reduction in apatite — combined influences of inherent instability, annealing anisotropy
length bias and system calibration. Length bias and system calibration. Earth Planet.
Sci. Lett. 89 (3–4), 335–352.
Green, P.F., Duddy, I.R., Hegarty, K.A., 2002. Quantifying exhumation from apatite fissiontrack analysis and vitrinite reflectance data: precision, accuracy and latest results
from the Atlantic margin of NW Europe. Geol. Soc. Lond. Spec. Publ. 196 (1), 331–354.
Green, P.F., Japsen, P., Chalmers, J.A., Bonow, J.M., 2011. Thermochronology, erosion surfaces and missing section in West Greenland. J. Geol. Soc. 168 (4), 817–829.
Green, P.F., Lidmar-Bergstrom, K., Japsen, P., Bonow, J.M., Chalmers, J.A., 2013. Stratigraphic landscape analysis, thermochronology and the episodic development of elevated,
passive continental margins. Geol. Surv. Den. Greenl. 30, 1–150.
Hendriks, B., 2003. Cooling and denudation of the Norwegian and Barents Sea margins,
Northern Scandinavia. Constrained by Apatite Fission Track and (U–Th)/He
Thermochronology. Vrije Universiteit, Amsterdam ((PhD), 177 pp.).
Hendriks, B., Andriessen, P., Huigen, Y., Leighton, C., Redfield, T., Murrell, G., Gallagher, K.,
Nielsen, S.B., 2007. A fission track data compilation for Fennoscandia. Nor. J. Geol. 87
(1–2), 143–155.
Hendriks, B.W.H., Osmundsen, P.T., Redfield, T.F., 2010. Normal faulting and block tilting
in Lofoten and Vesteralen constrained by apatite fission track data. Tectonophysics
485 (1–4), 154–163.
Japsen, P., Chalmers, J.A., Green, P.F., Bonow, J.M., 2012. Elevated, passive continental margins: not rift shoulders, but expressions of episodic, post-rift burial and exhumation.
Glob. Planet. Chang. 90–91, 73–86.
Japsen, P., Green, P.F., Chalmers, J.A., 2013. The mountains of North-East Greenland are
not remnants of the Caledonian topography. A comment on Pedersen et al. (2012).
Tectonophysics 589, 234–238.
Ksienzyk, A.K., Dunkl, I., Jacobs, D., Fossen, H., Kohlmann, F., 2014. From orogen to passive
margin: constraints from fission track and (U–Th)/He analyses on Mesozoic uplift
and fault reactivation in SW Norway Geological Society. Lond. Spec. Publ. 390,
689–702.
Kwon, Y.W., Bang, H., 2000. The finite element method using MATLAB. CRC Mechanical
Engineering Series. CRC Press, Boca Raton, FL.
Lidmar-Bergstrom, K., Ollier, C.D., Sulebak, J.R., 2000. Landforms and uplift history of
southern Norway. Glob. Planet. Chang. 24 (3–4), 211–231.
Lidmar-Bergstrom, K., Naslund, J.O., Ebert, K., Neubeck, T., Bonow, J.M., 2007. Cenozoic
landscape development on the passive margin of northern Scandinavia. Nor. J. Geol.
87 (1–2), 181–196.
Lidmar-Bergstrom, K., Bonow, J.M., Japsen, P., 2013. Stratigraphic landscape analysis and
geomorphological paradigms: Scandinavia as an example of Phanerozoic uplift and
subsidence. Glob. Planet. Chang. 100, 153–171.
Medvedev, S., Hartz, E.H., Podladchikov, Y.Y., 2008. Vertical motions of the fjord regions of
central East Greenland: impact of glacial erosion, deposition, and isostasy. Geology 36
(7), 539–542.
Medvedev, S., Souche, A., Hartz, E.H., 2013. Influence of ice sheet and glacial erosion on
passive margins of Greenland. Geomorphology 193, 34–46.
Molnar, P., England, P., 1990. Late Cenozoic uplift of mountain-ranges and global climate
change — chicken or egg. Nature 346 (6279), 29–34.
Nielsen, S.B., Gallagher, K., Leighton, C., Balling, N., Svenningsen, L., Jacobsen, B.H.,
Thomsen, E., Nielsen, O.B., Heilmann-Clausen, C., Egholm, D.L., Summerfield, M.A.,
Clausen, O.R., Piotrowski, J.A., Thorsen, M.R., Huuse, M., Abrahamsen, N., King, C.,
Lykke-Andersen, H., 2009. The evolution of western Scandinavian topography: a review of Neogene uplift versus the ICE (isostasy–climate–erosion) hypothesis.
J. Geodyn. 47 (2–3), 72–95.
Nielsen, S.B., Clausen, O.R., Jacobsen, B.H., Thomsen, E., Huuse, M., Gallagher, K., Balling, N.,
Egholm, D., 2010a. The ICE hypothesis stands: how the dogma of late Cenozoic
245
tectonic uplift can no longer be sustained in the light of data and physical laws.
J. Geodyn. 50 (2), 102–111.
Nielsen, S.B., Clausen, O.R., Pedersen, V.K., Leseman, J.E., Goledowski, B., Huuse, M.,
Gallagher, K., Summerfield, M.A., 2010b. Discussion of Gabrielsen et al. (2010): latest
Caledonian to present tectonomorphological development of southern Norway. Mar.
Pet. Geol. 27 (6), 1285–1289.
Olesen, O., Kierulf, H.P., Bronner, M., Dalsegg, E., Fredin, O., Solbakk, T., 2013. Deep
weathering, neotectonics and strandflat formation in Nordland, northern Norway.
Nor. J. Geol. 93 (3–4), 189–213.
Osmundsen, P.T., Redfield, T.F., Hendriks, B.H.W., Bergh, S., Hansen, J.A., Henderson, I.H.C.,
Dehls, J., Lauknes, T.R., Larsen, Y., Anda, E., Davidsen, B., 2010. Fault-controlled alpine
topography in Norway. J. Geol. Soc. 167 (1), 83–98.
Pedersen, V.K., Nielsen, S.B., Gallagher, K., 2012. The post-orogenic evolution of the Northeast Greenland Caledonides constrained from apatite fission track analysis and inverse geodynamic modelling. Tectonophysics 530, 318–330.
Pedersen, V.K., Nielsen, S.B., Gallagher, K., 2013. Reply to: “The mountains of North-East
Greenland are not remnants of the Caledonian topography. A comment on Pedersen
et al. (2012)”. Tectonophysics 589, 239–244.
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 32 (2), 161–164.
Peltier, W.R., 2004. Global glacial isostasy and the surface of the ice-age earth: the ice-5G
(VM2) model and grace. Annu. Rev. Earth Planet. Sci. 32, 111–149.
Perez-Gussinye, M., Watts, A.B., 2005. The long-term strength of Europe and its implications for plate-forming processes. Nature 436 (7049), 381–384.
Redfield, T.F., 2010. On apatite fission track dating and the Tertiary evolution of West
Greenland topography. J. Geol. Soc. 167 (2), 261–271.
Redfield, T.F., Osmundsen, P.T., 2009. The Jellefonna fault system of Western Norway:
linking late-Caledonian extension, post-Caledonian normal faulting, and Tertiary
rock column uplift with the landslide-generated tsunami event of 1756.
Tectonophysics 474 (1–2), 106–123.
Redfield, T.F., Osmundsen, P.T., 2013. The long-term topographic response of a continent
adjacent to a hyperextended margin: a case study from Scandinavia. Geol. Soc. Am.
Bull. 125 (1–2), 184–200.
Redfield, T.F., Torsvik, T.H., Andriessen, P.A.M., Gabrielsen, R.H., 2004. Mesozoic and Cenozoic tectonics of the More Trondelag Fault Complex, central Norway: constraints from
new apatite fission track data. Phys. Chem. Earth 29 (10), 673–682.
Redfield, T.F., Braathen, A., Gabrielsen, R.H., Osmundsen, P.T., Torsvik, T.H., Andriessen,
P.A.M., 2005a. Late mesozoic to early Cenozoic components of vertical separation
across the More–Trondelag Fault Complex, Norway. Tectonophysics 395 (3–4),
233–249.
Redfield, T.F., Osmundsen, P.T., Hendriks, B.W.H., 2005b. The role of fault reactivation and
growth in the uplift of western Fennoscandia. J. Geol. Soc. 162, 1013–1030.
Reusch, H., 1901. Nogle bidrag til forstaaelsen af hvorledes Norges dale og fjelde er blevne
til. Nor. Geol. Unders. 32, 124–263.
Riis, F., 1996. Quantification of Cenozoic vertical movements of Scandinavia by correlation
of morphological surfaces with offshore data. Glob. Planet. Chang. 12 (1–4), 331–357.
Slagstad, T., Balling, N., Elvebakk, H., Midttomme, K., Olesen, O., Olsen, L., Pascal, C., 2009.
Heat-flow measurements in Late Palaeoproterozoic to Permian geological provinces
in south and central Norway and a new heat-flow map of Fennoscandia and the
Norwegian-Greenland Sea. Tectonophysics 473 (3–4), 341–361.
Steer, P., Huismans, R.S., Valla, P.G., Gac, S., Herman, F., 2012. Bimodal Plio-Quaternary glacial erosion of fjords and low-relief surfaces in Scandinavia. Nat. Geosci. 5 (9),
635–639.
Stern, T.A., Baxter, A.K., Barrett, P.J., 2005. Isostatic rebound due to glacial erosion within
the Transantarctic Mountains. Geology 33 (3), 221–224.
Torsvik, T.H., Smethurst, M.A., Burke, K., Steinberger, B., 2008. Long term stability in deep
mantle structure: evidence from the similar to 300 Ma Skagerrak-Centered Large Igneous Province (the SCLIP). Earth Planet. Sci. Lett. 267 (3–4), 444–452.
Tsikalas, F., Eldholm, O., Faleide, J.I., 2005. Crustal structure of the Lofoten–Vesterdlen continental margin, off Norway. Tectonophysics 404 (3–4), 151–174.
Turcotte, D.L., Schubert, G., 2002. Geodynamics. Cambridge University Press, Cambridge,
New York.
Watts, A.B., 2001. Isostasy and Flexure of the Lithosphere. Cambridge University Press,
Cambridge; New York.
Weisz, G., 1992. An investigation of Jurassic coals from Haltenbanken and Beistadfjorden.
A comparison of Composition and Maturity. Technical University of Trondheim
(NTNU), Trondheim, Norway ((Diploma thesis), 72 pp.).