high-resolution dtm-based stratigraphic correlation of
TRANSCRIPT
320 Kartographische Nachrichten 6.2017
GIS IN GEOMORPHOLOGY | T. Bueche, K. Loesch: Stratigraphic correlation of fluvial terraces
Brückner 1901/09). The River Würm inter
sects the Würmian terminal moraines to the
north of the lake. Along this section several
fluvial terraces with various levels of eleva
tion exist indicating different stages of in
cision (Schulz 1989). The transverse valley
of River Würm has been mapped geomor
phologically by Schulz (1989), who outlines
all fluvial terraces in one class. Jerz (1987)
classified the terraces to be either Plenigla
cial Würmian or Late Würmian. No further
detailed stratigraphic differentiation of the
fluvial terraces along River Würm exists.
In this work, we present the identification
of fluvial terraces using a semiautomated
GIS approach. Terraces were indicated for a
terraced section of River Würm from a DTM
with high spatial resolution of 1 m. The
approach of terrace derivation from a DTM
is frequently used and several methods exist
(Hopkins & Snyder 2016) such as the fre
quently used semiautomated TerEx Toolbox
timeconsuming and requires financial re
sources (Jones et al. 2007; Stout & Belmont
2014). Further constraints can be obstructed
lines of sight, restrictions on land access, or
vegetation cover. Digital terrain models
(DTM) with high resolutions are ideally
suited for a systematic mapping of geomor
phological structures with high accuracy
(Passalacqua et al. 2012; Beckenbach et al.
2014; Dühnforth et al. 2017) and for its
geological interpretation. This detailed spa
tial information enables the creation of
maps indicating fluvial terraces related to
the same period of origin, which enhances
the understanding of the genesis of a river
valley and its various stages of incision.
The area around the Starnberger See (former
called Würmsee) on the Alpine foreland is
of great geomorphological interest as the
sequence of glacial deposits there is a type
section of the Würmian glaciation (Reuther
et al. 2011) and eponymous for it (Penck &
1 Introduction
River terraces are fundamental archives of
the landscape evolution (Kolb et al. 2016)
and their identification and localization is a
classic part of geomorphological mapping
(e. g. von Asselen & Seijmonsbergen 2006;
Stahl et al. 2013). Fluvial terraces provide
information about the geomorphological
genesis of valleys and the timing of the
river incision (Demoulin et al. 2007; Kolb et
al. 2016; Limaye & Lamb 2016) and hence
are also indicators for the environmental
conditions at the time of terrace formation
(Pavlopoulos et al. 2009; Fryirs & Brierley
2012). To be able to distinguish different
stages of the incision it is important to cor
relate terraces, treads, or steps at various
river segments chronostratigraphically. A
traditional geomorphological mapping and
groundbased measuring of fluvial terraces
for several sites and river segments is
High-resolution DTM-based stratigraphic correlation of fluvial terraces along River WürmStratigraphische Zuordnung von Flussterrassen entlang der Würm unter Verwendung eines hochaufgelösten Digitalen Geländemodells
Thomas Bueche, Kilian Loesch; München
Digital terrain models (DTM) with high resolution are ideally suited for a systematic mapping of geomorpholog-
ical structures, such as fluvial terraces, with high accuracy. We developed a semi-automatic GIS method to correlate
individual terrace treads stratigraphically using a 1 m DTM. It is applied to a gorge of River Würm (Southern
Bavaria, Germany). The approach yields suitable and plausible results and the presented map provides a clear and
easy overview of correlated terrace levels. The investigation enhances the understanding of the spatial and tem-
poral evolution of the valley and demonstrates the great benefit of high-resolution DTMs and the digital process-
ing of geodata for the geomorphological research.
n Keywords: High-resolution DTM, geomorphological mapping, fluvial terraces, geodata processing,
landscape evolution, River Würm
Hochaufgelöste Digitale Geländemodelle (DGM) sind bestens geeignet, um geomorphologische Strukturen wie Flussterrassen systematisch und mit hoher Genauigkeit zu kartieren. Wir haben einen semi-automatischen GIS-Ansatz entwickelt, der unter Verwendung eines 1-m-DGM die Ableitung von stratigraphisch zusammengehören-den Terrassenflächen ermöglicht. Die Methode wurde auf einen Abschnitt der Würm (Südbayern) angewendet und liefert geeignete und plausible Ergebnisse. Die resultierende Karte gibt einen strukturierten Überblick zur strati-graphischen Zuordnung der Terrassen. Die Untersuchung verbessert das Verständnis zur räumlichen und zeitlichen Genese des Tals und zeigt das große Potenzial von hochaufgelösten DGM in der digitalen Geodatenverarbeitung und für die geomorphologische Forschung.n Schlüsselwörter: Hochaufgelöstes DGM, geomorphologische Kartierung, Flussterrassen,
Geodatenverarbeitung, Landschaftsgenese, Würm
321Kartographische Nachrichten 6.2017
T. Bueche, K. Loesch: Stratigraphic correlation of fluvial terraces | GIS IN GEOMORPHOLOGY
(Nicoulin 2014; Stout & Belmont 2014), the
Rahnis method suggested by Walter et al.
(2007), and the method presented by Clubb
et al. (2017). Within these approaches, either
the step of a stratigraphic correlation of
terraces of the same incision stage along
different river segments is not included
(TerEx, Clubb et al. method), or previous
knowledge about the correlation of terraces
is required (Rahnis Method). Therefore, we
focus on the processing step of extrapolat
ing terrace surfaces to “infinity” in order to
correlate them with terraces at other sections
of the river without any requirements of
prior knowledge about the stratigraphic
genesis of the area. Three methods to ex
trapolate terrace surfaces are presented and
tested. The presentation of the results of the
fluvial terrace surface extrapolation is not
only a mere visualization of the terraces
(e. g. provided by TerrainView, Beckenbach
et al. 2014) but the mapping also displays
the identified stratigraphically correlated
treads along River Würm.
2 Study area
The River Würm intersects the ridges of the
terminal moraines of Würmian last glacial
maximum (LGM, Pleniglacial, Doppler et al.
2011) and stages of deglaciation of this last
glaciation between the settlements Leutstet
ten and Gauting (see Fig. 1). The gorge was
the main meltwater channel during the last
glaciation (Jerz 1987) and the river still
drains the Starnberger See, which fills the
overdeepened basin eroded by a lobe of the
IsarLoisach Glacier during the Pleistocene
(Preusser et al. 2010; Reuther et al. 2011).
This channel is also incised deeply into the
main Würmian alluvial fan called ‘Münch
ner Schotterebene’, which is adjacent to the
north of the moraine ridges and is dated to
be pleniglacial origin and can be also con
sidered as terrace (Doppler et al. 2011). In
addition, older moraine and gravel deposits
of the Rissian glaciation outcrop in this area.
In the study area (2 x 2 km square section)
numerous fluvial terraces exist along River
Würm (Fig. 2). The area covers a meander of
the river, where several treads divided by
sharp erosion edges have developed. The
extrapolation of the terrain is realized for
each terrace surface part of this “stairway”
of several treads (black outlined area in
Fig. 2). The geomorphologic units and ero
sion edges shown in Fig. 2 are derived from
the Geological Map Bavaria (scale 1:25000,
GK25, Bavarian Environment Agency, www.
lfu.bayern.de). The present level of the river
has an elevation of 582.3 m a.s.l. at the
southern margin of the study area, where the
peaks of the adjacent moraines are elevated
above with an elevation of 632.0 m a.s.l. The
River Würm leaves the study area in the
north at a height of 567.0 m a.s.l., incised by
35 m (to the left) and 26 m (to the right) into
the main Würmian Pleniglacial terrace.
3 Data and Methods
3.1 Data, software and
stratigraphic nomenclature
The analysis is based on a DTM with a spa
tial resolution (pixel size) of 1 x 1 m and an
elevation accuracy of < 0.1 m. The DTM is
a product of processed laser scan data pro
vided by the Bavarian State Office for Sur
vey and Geoinformation. All geodata pro
cessing is performed with ESRI ArcGIS 10.3
and specific tools used for the analyses are
implemented in the ArcToolbox provided by
ArcGIS. The nomenclature of the strati
graphic system used in this study refers to
Doppler et al. (2011).
3.2 Identification of terraces
and extrapolation approach
In a first step, relevant fluvial terraces with
in the area of interest (see black outlined
polygon in Fig. 2) are identified. We define
a fluvial terrace as a platform with a weak
inclination of an average slope smaller than
Fig. 1: Overview on the geomor- phological setting in the area of Isar-Loisach Glacier (modiied after Rippl 2011 and Kleinmann 1995)
2°, which is delimited by terrain break lines.
Although the geological mapping (1:25.000)
includes erosional edges shown in Fig. 2,
these data are not provided in a sufficient
resolution. Therefore, we delineate the rele
vant terrain break lines from the DTM
manually. As they describe a discontinuity
of the local surface (Briese et al. 2009), these
line features trace highly inclined areas (see
Fig. 3) and can be derived from the terrain’s
slope. Furthermore, the vertical curvature
(Spatial Analyst Tool vertical curvature) is
calculated from the DTM highlighting sharp
edges of the relief and it is also taken into
account during the digitalization process of
the terrain break lines. The selection of
terraces and the delineation of break lines
are restricted to areas below the uppermost
level of the main pleniglacial terrace plain.
The final assessment of the terraces is based
on expert knowledge considering the slope
and break lines.
In total, seven terrace polygons are iden
tified to be part of the stairway at the
river meander (Fig. 3). These terraces are
chosen to be extrapolated to the extent of
the entire study area to correlate them
stratigraphically with other fluvial terraces.
All seven terraces are outlined by break
lines of the relief. Two cross sections of the
treads (A and B, see Fig. 3) visualize the
vertical and horizontal offsets between the
terraces.
322 Kartographische Nachrichten 6.2017
GIS IN GEOMORPHOLOGY | T. Bueche, K. Loesch: Stratigraphic correlation of fluvial terraces
The process of extrapolation of the fluvial
terraces to “infinity” is executed by gener
ating an (inclined) plane representing a
terrace surface. Four specifications are re
quired to define such a plane using vector
geometry:
• the coordinates (x and yvalues) of the
terrace center,
• the mean elevation height at the terrace
center,
• the aspect (Spatial Analyst Tool Aspect)
of the surface,
• and its dip, i. e. the mean slope (Spatial
Analyst Tool Slope).
In addition, to create a data set for the study
area the elevation heights of the square
vertexes have to be calculated, since there
exists no tool in ArcGIS to calculate an
inclined plane from the specifications listed
above. The coordinates for the center points
of each terrace are calculated using the
respective function of ArcGIS (Calculate
Geometry). Mean heights of the terrace sur
faces, respectively, are obtained by the
Spatial Analyst Tool Zonal Statistics as
Tables. Three approaches are tested to deter
mine the aspect and dip of each terrace:
1. The calculation based on values for aspect
and slope for each pixel of the terrace:
After the generation of these values for
all pixels of the DTM the mean slope and
the major aspect for each terrace is ob
tained using the Spatial Analyst Tool
Zonal Statistics as Tables. For this calcu
lation, only pixels with a slope <= 3° are
considered.
2. The calculation based on random points
– individual for each terrace: A new
vector data set of random points (n =
1,000) is created for each terrace restric
ted by its polygon, respectively. The
elevation height for each point is extrac
ted from the DTM to the attribute table
of the Point FeatureClass using the
Spatial Analyst Tool Extract Values to
Points. Based on this elevation informa
tion an inclined surface plain is interpo
lated applying the Spatial Analyst Tool
Trend. Subsequently, the aspect and
slope of this surface plain can be calcu
lated.
3. The calculation based on random points
– general for the study area: Instead of
calculating the aspect and dip for each
terrace, respectively, the general values
of those two required specifications are
determined for the entire study area as
suming to be representative for all terra
ce levels. For that purpose, a new vector
data set of random points (n = 50,000) is
created for the study area. Out of this
sample, points are selected (Select by
Location) for areas with a slope of <= 3°
to discard erosional edges. Subsequently,
the general aspect and slope of the enti
re study area is calculated following the
same steps as described for method 2.
Once the four specifications of a terrace
surface are obtained, it can be extrapolated
as an inclined plane to the whole study
area by calculating the elevation heights
(zvalues) of the four square vertexes of the
2 x 2 km polygon. A 3DPolygon is created
using the 3D Analyst Tool 3D-ASCII in
Feature-Class, specifying the x, y, and
zvalue.
The application of method 1 does not yield
a realistic distribution of pixel values for the
aspect. All values with a magnification of
45° (e. g. 0°, 45°, 90°, …) exceed any other
value approximately by the factor 2. As a
consequence, only these values are obtained
as major aspect for every terrace and, hence,
this method is discarded and not applied. An
explanation for this result could be the
preprocessing of the DTM data, when an
automated filter eliminates errors of the raw
laser scan data such as houses or vegetation
(Mr Boxler 2017, pers. communication,
8 August). Method 2 describes the individ
ual specifications of the aspect and dip of
the terraces very locally. This is suitable to
detect the correlation of terraces which are
in a short distance to the recent river. With
increasing heights of fluvial terraces above
the valley bottom, and accordingly increas
ing age, the terrain characteristics become
less dependent from the recent river chan
nel, but more represented by the wider en
vironment. Thus, method 2 (individual) is
found to be applicable only for the terrace
of lowest elevation (7) (Fig. 3), as the diver
gence of the aspect of the valley segment to
that of the general terrain is significant here.
For terraces 1–6 (Fig. 3), method 3 (general)
is applied.
3.3 Stratigraphic correlation of terraces
Fluvial terraces along the river are de
termined to correlate stratigraphically
when the difference of the elevation height
Fig. 2: Shaded relief derived from the 1 m resolution DTM (Geobasisdaten © Bavarian State Oice for Survey and Geoinformation) of the study area with geological units and erosion structures from the Geological Map Bavaria (scale 1:25000, GK25, Bavarian Environment Agency, www.lfu.bayern.de, map sheet Starnberg-Nord 7934). The area of interest containing the "stairway” of several luvial terraces is displayed by the black outlined polygon. The anthropogenic modiied areas do not include the gravel pits and streets due to the reduced resolution of the geological units
323Kartographische Nachrichten 6.2017
T. Bueche, K. Loesch: Stratigraphic correlation of fluvial terraces | GIS IN GEOMORPHOLOGY
identified as correlated terraces even if they
are not completely delineated by break
lines.
4 Results
Fig. 5 shows the result of the stratigraphic
correlation of the fluvial terrace levels in the
study area. Additional to the seven terrace
levels, “interstages” are determined, when
terraces are intersected by the extrapolation
of two terrace levels. Those terrace levels,
which are not identified to be within a
1.5 m difference to an extrapolated terrace
surface, are correlated as an interstage to the
closest level (e. g. the yellowhatched terrace
in the northwest, which is accordingly in
dicated close to Terrace 3). The map shows
that several terraces could be correlated
directly to the extrapolated levels of the
stairway at the river meander. The existence
of a couple of interstages indicates an epi
sodic downcutting of the valley in many
small stages (Pierce et al. 2011). All older
terraces are located at the valley sides,
whereas lower surface levels partly retained
as older platforms above surrounded com
pletely by younger terraces. The extent of
the two highest levels of earliest develop
ment (Terraces 1 and 2) cover more than the
half of the area, which is incised into the
main Würmian pleniglacial terrace down
stream of the river meander. This suggests
surface of terrace 2 and the correlated are
as (marked in blue). The final assessment of
correlated areas is only realized (see. Fig. 5)
for those, which meet the specifications for
terraces defined above. Intersections of the
extrapolated terrace with the terrain at
highly inclined areas do not represent flu
vial terraces (e. g. blue line on right side
(west) of the valley, Fig. 4) and can hence
be discarded. Nevertheless, areas are also
to the extrapolated plain is smaller than 1.5
m. Thus, the difference in height is calcu
lated for each pixel of the DTM by convert
ing the output 3D vector data set of the
extrapolation process in a raster surface (3D
Analyst Tool Minus). Several thresholds for
the maximum difference between the ex
trapolated surface and terrain are tested and
1.5 m is found to yield the most realistic
value. Figure 4 visualizes the extrapolated
Fig. 3: Areas of the luvial terraces of the examined stairway are shown in coloured polygons. The background of the map igures the terrain’s slope displaying high values in dark shades. Two proile lines (A, C-D-E), orange; B, C-D-F, blue) trace the elevation of the steps of the terrain
Fig. 4: 3D visualization of the study area. The elevation (gray shades – bright: high, dark: low) is based on the 1 m DTM pictured with a vertical exaggeration by the factor of 7. The transparent yellow plain represents the extrapolated surface of terrace 2 (outlined in red). All blue areas marking pixels of the DTM within 1.5 m diference of height to the extrapolated inclined surface plain
324 Kartographische Nachrichten 6.2017
GIS IN GEOMORPHOLOGY | T. Bueche, K. Loesch: Stratigraphic correlation of fluvial terraces
gated river section is not mainly straight
but partly sinuously. This problem could
be addressed by the flexible application
of two methods to deduce the specifications
of dip and aspect for the surface extrapola
tion. The choice of the suitable method for
each terrace, respectively, can hardly be
automated and requires an individual deci
sion.
Beside the required expert knowledge, fur
ther limitations of this approach have to be
mentioned. The extrapolation approach
includes the assumption of a linear slope of
the terrain. As this simplification of the real
conditions does not trace any changes of
river channel slope, it induces errors for the
correlation of terrace surfaces. This effect
will intensify with increasing distance to the
extrapolated terrace surface. Hence, the
method is only suitable for examinations at
smaller scales.
The applied threshold for the maximum
vertical deviation of 1.5 m to correlate ter
race levels with the extrapolated treads is
found to yield realistic results. It is sitespe
cific and should be in the range from 50 to
75 % of the minimum separation in height
between the examined terraces (here: 2.2 m,
Fig. 3) to avoid too many correlations of
terraces to two treads on the one hand, and
terraces without intersection with any ex
trapolated terrace surface.
The presented approach differs from the
Rahnis method (Walter et al. 2007 in Hop
kins & Snyder 2016), which extrapolates
fluvial terraces to 3D surfaces based on
points manually set on several different
terrace treads. This approach enables a cor
relation of terraces of a broader environ
ment, but requires already previous knowl
edge of potential stratigraphic correlation of
terraces. As our method attains the correla
tion of individual terrace treads, it provides
the required information to be able to apply
the Rahnis method.
Despite the identified limitations, the meth
od can be evaluated as very practicable. The
obtained results provide a very detailed map,
which enables an interpretation of the gen
esis of the river channel. Due to the options
of different methods to derive the specifica
tions of dip and aspect for the surface ex
trapolation depending on the distance to the
recent river, the approach can be easily
transferred to other sites with complex
courses of valleys.
geodata processing, the analysis could be
performed without any groundbased meas
urements. However, this approach requires
expert knowledge and includes a process of
manual digitalization of terrain break lines.
Thus, it can only be characterized as a
semiautomated approach. Although auto
matic approaches exist to delineate break
lines from DTM data (e. g. Rutzinger et al.
2012), they hardly detect geomorphic struc
tures in continuous features, which is a re
quirement for those purposes.
All methods presented in this work to deter
mine the aspect and dip of terrace surfaces
include the specification of a threshold for
the maximum slope by the user. The herein
applied limit of 3° can be seen as a suitable
threshold to discard hillslopes and scarps.
Alterations of this slope threshold will have
an impact on the final identification of
terrace surfaces.
The process of terrace extrapolation is
complicated by the fact that the investi
the existence of a broad river valley during
the initial deepening in the Pleniglacial.
Furthermore, considering the geological
units of the geological mapping (Fig. 2),
it can be deduced that the evolution of
the shape of the valley to a narrow
gorge already took place before the Late
Glaciation period accompanied by progres
sive incision.
5 Discussion
The presented approach of extrapolating
terrace surfaces to correlate them with oth
er terraces along a river yields suitable and
plausible results, which does not require any
previous knowledge of tread correlations.
The method enables a very detailed mapping
of the terraced valley providing a clear and
easy overview of the correlated terrace
levels, which is an improvement to the vis
ualization tool TerrainView developed by
Beckenbach et al. (2014). By using solely
Fig. 5: Stratigraphic correlation of detected luvial terraces (labelled polygons) with other terrace levels in the study area. A correlation of an area is depicted by the same symbology as the extrapolated terrace. Hatched areas of two colours mark levels which were correlated to two terraces (interstages). Hatched terraces with only one colour show areas which were not within the height diference to be correlated. Their colour indicate the terrace with the smallest diference of height
325Kartographische Nachrichten 6.2017
T. Bueche, K. Loesch: Stratigraphic correlation of fluvial terraces | GIS IN GEOMORPHOLOGY
Preusser, F.; Reitner, J. M.; Schlüchter, C. (2010): Distribution, geometry, age and origin of overdeepened valleys and basins in the Alps and their foreland. Swiss Journal of Geosciences, 103 (3): 407–426
Reuther, A. U.; Fiebig, M.; IvyOchs, S.; Kubik, P. W.; Reitner, J. M.; Jerz, H., Heine, K. (2011): Deglaciation of a large piedmont lobe glacier in comparison with a small mountain glacier – new insight from surface exposure dating. Two studies from SE Germany. Quaternary Science Journal, 60 (23): 248–269. doi: 10.3285/eg.60.23.03
Rippl, C. (2011): Thermische Entwicklung des Ammersees als regionale Auswirkung des Globalen Wandels. LudwigMaximiliansUniversity, Munich
Rutzinger, M.; Höfle, B.; Kringer, K. (2012): Accuracy of automatically extracted geomorphological breaklines from airborne LiDAR curvature images. Geografiska Annaler: Series A, Physical Geography, 94 (1): 33–42. doi:10.1111/j.14680459.2012.00453.x
Schulz, G. (1989): Lexikon zur Bestimmung der Geländeformen in Karten, vol 28. Univerlag TU Berlin
Stahl, T.; Winkler, S.; Quigley, M.; Bebbington, M.; Duffy, B., Duke, D. (2013): Schmidt hammer exposure‐age dating (SHD) of late quaternary fluvial terraces in New Zealand. Earth Surface Processes and Landforms, 38 (15): 1838–1850
Stout, J. C.; Belmont, P. (2014): TerEx Toolbox for semiautomated selection of fluvial terrace and floodplain features from lidar. Earth Surface Processes and Landforms, 39 (5): 569–580. doi:10.1002/esp.3464
von Asselen, S.; Seijmonsbergen, A. C. (2006): Expertdriven semiautomated geomorphological mapping for a mountainous area using a laser DTM. Geomorphology, 78: 309–320
Walter, R.; Merritts, D.; Rahnis, M. (2007): Estimating the Volume, Nutrient Content, and Rates of Stream Bank Erosion of Legacy Sediment in the Piedmont and Valley and Ridge Physiographic Provinces, Southeastern and Central PA. Report. Pennsylvania Department of Environmental Protection, Harrisburg, Pennsylvania
About the authors
Dr. Thomas Büche is a postdoctoral researcher at the Department of Geography of LudwigMaximiliansUniversity Munich. His research includes GISbased applications for analyses and visualization of the landscape topography, and the impact of climate and land use change on lake systems. His email: [email protected].
M. Sc. Kilian Loesch holds a degree in Environmental Systems and Sustainability from the Department of Geography at the LudwigMaximiliansUniver sity in Munich. His research interests include GISbased and modelling techniques to analyze environmental systems. A particular focus is set on analyses of environmental systems affected by climate change. His email: [email protected].
Manuscript submitted: 20171003 Accepted after review: 20171031
Doppler, G.; Kroemer, E.; Rögner, K.; Wallner, J.; Jerz, H.; Grottenthaler, W. (2011): Quaternary stratigraphy of southern Bavaria. Quaternary Science Journal, 60 (23): 329–365
Dühnforth, M.; Densmore, A. L.; IvyOchs, S.; Allen, P.; Kubik, P. W. (2017): Early to Late Pleistocene history of debrisflow fan evolution in western Death Valley (California) using cosmogenic 10 Be and 26 Al. Geomorphology, 281: 53–65. doi:10.1016/ j.geomorph.2016.12.020
Fryirs, K. A.; Brierley, G. J. (2012): Geomorphic Analysis of River Systems: An Approach to Reading the Landscape. In: Geomorphic Analysis of River Systems. John Wiley & Sons, Ltd, pp 1–8. doi: 10.1002/9781118305454.ch1
Hopkins, A. J.; Snyder, N. P. (2016): Performance evaluation of three DEMbased fluvial terrace mapping methods. Earth Surface Processes and Landforms, 41 (8): 1144–1152. doi:10.1002/esp.3922
Jerz, H. (1987): Geologische Karte von Bayern 1:25000, Erläuterungen zum Blatt Nr. 7934 Starnberg Nord. München
Jones, A. F.; Brewer, P. A.; Johnstone, E.; Macklin, M. G. (2007): Highresolution interpretative geomorphological mapping of river valley environments using airborne LiDAR data. Earth Surface Processes and Landforms, 32 (10): 1574–1592. doi:10.1002/esp.1505
Kleinmann, A. (1995): Seespiegelschwankungen am Ammersee. Ein Beitrag zur spätund postglazialen Klimageschichte Bayerns. Geologica Bavarica, 99: 253–367
Kolb, T.; Fuchs, M.; Zöller, L. (2016): Deciphering fluvial landscape evolution by luminescence dating of river terrace formation: a case study from Northern Bavaria, Germany. Zeitschrift für Geomorphologie, Supplementary Issues, 60 (1): 29–48. doi:10.1127/zfg_suppl/2015/S00193
Limaye, A.; Lamb, M. P. (2016): Numerical model predictions of autogenic fluvial terraces and comparison to climate change expectations. Journal of Geophysical Research: Earth Surface:
Nicoulin, A. (2014): Fluvial Terraces and PostGlacial River Incision Along the Farmington and Housatonic Rivers in Southern New England. Masters, University of Connecticut
Passalacqua, P.; Belmont, P.; FoufoulaGeorgiou, E. (2012): Automatic geomorphic feature extraction from lidar in flat and engineered landscapes. Water Resources Research, 48 (3): n/an/a. doi:10.1029/ 2011WR010958
Pavlopoulos, K.; Evelpidou, N.; Vassilopoulos, A. (2009): MethodologyTechniques. In: Mapping Geomorphological Environments. Springer Berlin Heildelberg, pp 5–47. doi:10.1007/9783642019500_1
Penck, A.; Brückner, E. (1901/09): Die Alpen im Eiszeitalter, vol 3. Tauchnitz Leipzig
Pierce, J. L.; Meyer, G. A.; Rittenour, T. (2011): The relation of Holocene fluvial terraces to changes in climate and sediment supply, South Fork Payette River, Idaho. Quaternary Science Reviews, 30 (5): 628–645
6 Conclusion
In this study a detailed stratigraphic corre
lation of fluvial terraces along a section of
River Würm is presented based on DTM data
only, which requires no previous knowledge
of terrace level correlations. Beyond the
delineation of fluvial terraces, the presented
approach provides an analysis of the corre
lation of individual fluvial terraces at dif
ferent river sections, which can be trans
ferred to other areas. It is shown that the
method is applicable to river sections cov
ering meanders. The output results in a first
geomorphological mapping of the fluvial
terraces for the study area and benefits to a
better understanding of the spatial evolution
of this transverse valley. The investigation
demonstrates the very suitable applicability
of highresolution DTM processed with GIS
tools, without any survey in the field or
groundbased measuring, for the detailed
mapping and visualization of geomorpho
logical structures and its geological inter
pretation.
Acknowledgements
We thank for the financial support of the
study by the teaching and research pro
gramme Lehre@LMU sponsored by the
German Federal Ministry of Education and
Research (BMBF) (grant no 01PL17016)
providing funding for the DTM data. We
also thank two anonymous reviewers pro
viding constructive comments that im
proved the manuscript.
Literature
Beckenbach, E.; Müller, T.; Seyfried, H., Simon, T. (2014): Potential of a highresolution DTM with large spatial coverage for visualization, identification and interpretation of young (Würmian) glacial geomorphology: a case study from Oberschwaben (southern Germany). Quaternary Science Journal, 63: 107129. doi:10.3285/eg.63.2.01
Briese, C.; Mandlburger, G.; Ressl, C., Brockmann, H. (2009): Automatic break line determination for the generation of a DTM along the river Main. Paper presented at the Laser scanning 2009, IAPRS, Paris
Clubb, F. J.; Mudd, S. M.; Milodowski, D. T.; Valters, D. A.; Slater, L. J.; Hurst, M. D., Limaye, A. B. (2017): Geomorphometric delineation of floodplains and terraces from objectively defined topographic thresholds. Earth Surface Dynamics, 5 (3): 369–385. doi:10.5194/esurf53692017
Demoulin, A.; Bovy, B.; Rixhon, G.; Cornet, Y. (2007): An automated method to extract fluvial terraces from digital elevation models: The Vesdre valley, a case study in eastern Belgium. Geomorphology, 91 (1): 51–64