1 Digital Identification and Mapping of Geomorphology in the Baruther-Urstromtal (Germany) using LiDAR data and other digital sources.
Bachelor Thesis Future Planet Studies, Major Earth Sciences, University of Amsterdam
Joost Bakker 30-5-2021 Amsterdam
Supervised by: dr. W.M. de Boer
2
Abstract
Het doel van dit onderzoek is het in kaart brengen van de geomorphologie van het Baruth-oerstroomdal. Het Baruth-oerstroomdal is gelegen in Duitsland ten zuiden van Berlijn en gevormd in de laatste 2 ijstijden. De geomorphologische kaart is gemaakt doormiddel van digitale bronnen zoals LiDAR data, luchtfoto’s en eerder gemaakte kaarten. Hierbij zijn verschillende landvormen in kaart gebracht waaronder morenen, verschillende
3
Contents
Abstract 2
Contents 3
Introduction 3
Methods and Data 5
Results 7 Discussion 14 Conclusion 17 Literature 17 Acknowledgements 18 Appendices 19
Introduction
Technological advancements that have taken place during the last few decades have led to a massive increase in efficiency of many aspects of human life. The scientific community and scientific research have greatly benefited from the abilities to share new information worldwide with a few tabs of a button. Also technological advancements have made it
possible to measure and map the world around us in an incredibly efficient and precise way. Traditionally geomorphological mapping was done through physical fieldwork which made the mapping of larger areas labour intensive and therefore expensive (Jones, Brewer, Johnstone, & Macklin, 2007). However, according to Geskus (2020) digitally mapping geomorphological features based on LiDAR data can, despite having some limitations, to some extent replace the traditional form of mapping through fieldwork.
Research Aims and Questions
In 2020 Geskus (2020), Nobel (2020), Luimes (2020), Romar (2020), Schadee (2020) and Zuidervaart (2020) created a geomorphological map on macro, meso and micro level of the Central Baruth Ice-Marginal Valley in Brandenburg Germany.
4 Figure 1: Research area 2020, 2021 and research strip.
This map was based on the legends of the mappings of Pachur & Schulz (198) and Frank (1987). The main goal of this research is to expand that map eastwards using the same legend and improving it where possible. This is done in collaboration with 3 research partners and by relying on solely digital available data like LiDAR data, existing maps and photos. In order to do this geomorphological features must be identified on macro, meso and micro level. This will also require the following research questions to be answered:
- What geomorphological features, landforms and structures can be identified using LiDAR data, aerial photos and other digital sources?
- To what extent have people influenced the geological features in the research area? - How far have the Ice masses in the last few Ice ages reached into the study area and how have they influenced the geological features of the area?
5
Methods and Data
Before the mapping preparations were made to be able to successfully identify and map the most important geomorphological features in the research area. First dr. W. M. de Boer, expert in the geomorphology of the study area, discussed and explained the most important and prevalent landforms in the study area and how they were formed. Afterwards more literature on the area has been studied, the most important of which are de Boer (1995), Juschus (2001), de Boer (2015), Geskus (2020), Nobel (2020), Luimes (2020), Romar (2020), Schadee (2020) and Zuidervaart (2020).
Thirdly, the e-learning courses from ESRI ‘Lidar in ArcGIS: an Introduction’, ‘Managing Lidar Data Using LAS Datasets’, ‘Managing Lidar Data Using Terrain Datasets’ and ‘Managing Lidar Data Using Mosaic Datasets’ had been completed to gain more knowledge about how to use the LiDAR data and the software that was required for this research.
After this the “How to build an APRX” file that was provided by supervisor dr. W.M. de Boer was used to create an ArcGIS Pro project file with the extension ‘.aprx’. The “How to build an APRX” file describes which projections to use, how to add the most important existing maps to the ‘.aprx’ file as well as how to turn the LiDAR data tiles into usable layers in the ‘.aprx’ file. The document also helped to create a similar basis for all of the research partners. Which made transferring data between the research partners a lot more efficient.
The most important data like the LiDAR tiles, previously made maps, photos and literature was provided by dr. W.M. de Boer and was shared through surfdrive.nl. For the sharing of data between the research partners Microsoft Teams was used. This was also used for discussions through video calls which will be mentioned later.
The 30 LiDAR tiles that have been used for this research consist of xyz ascii files and contain elevation data of millions of points forming a big point cloud of elevation data. These files are then transformed into .LAS files from which the digital elevation model has been generated. This DEM file has a value for every 0.5 meter and is therefore very accurate. which contain accurate data on the elevation of the research area were used to create a digital elevation model (DEM) of the area. This DEM could then be used to create different elevation derived maps such as a hillshade map, aspect map, contour map and slope map. These were then used together with the DEM, literature, existing maps, photos and satellite imagery to map the geomorphological features of the research area.
In order to locate and map important features of the area the dynamic range adjustment (DRA) was used on the DEM. This adjusts the symbology of the DEM layer to the values that are on the screen. This makes small, but sometimes important, differences in the elevation visible. Another tool to get a clearer image of features in the area is convolution which research partner van Gelderen (2021) has described in more detail.
6 Figure 2: Left: DRA off, Right: DRA on.
The legend of the geomorphological map will, just like the legend made by Schadee (2020), be based on the legend created by Frank (1987) and the legend created by Pachur &
Schultz (1983). According to Schadee the legend made by Frank (1987) is very detailed and it contains classifications of most landforms that can be found in Germany. Schadee (2020) also states that the legend created by Frank (1987) has been based on the legend made by Pachur & Schultz (1983) which was made for a similar area as the Central Baruther ice marginal valley. However, according to Schadee (2020) the existing legends needed some additions to better show the important features of the research area. Therefore the legend created by Schadee (2020) will be used for the geomorphological map of this year and identified features that are not yet in that legend will be added based on the legend created by Frank (1987).
In order to create the correct symbology to the geomorphological map Schadee (2020) added codes to the features. These codes have been based on the classifications of the legend. However the classifications in the legends use dots to differentiate between main classifications, landforms and categories. The codes used for the symbology in ArcGIS Pro can not contain “.”. Therefore the “.” has been replaced by “0”. This means for example that the legend unit 13.6 Fluvioglacial Accumulative will be marked with 1306 in the attribute table on ArcGIS Pro as can be seen in figure 3.
7 To better utilize the possibilities of digital mapping the different features of last year's feature class “Unit 13” of the macro geomorphology have been separated into different feature classes. The goal of this adjustment was to be able to overlap some features creating a more accurate representation of the geomorphology. Since the dunes and anthropogenic structures lay on top of other geomorphological features such as ground moraines this was the most accurate way to map these features. This also makes it possible to display only certain features, for example showing just the ground moraines without the dunes on top. According to Jones et al. (2007) the mapping of geomorphological features is often
subjected to interpretations and can be subjective. Therefore most of the mapped features during this research have been discussed with multiple research partners.
Results
This paragraph has been divided into 4 parts of which the first 3 will subsequently focus on the macro, meso and micro structures that have been mapped in the fieldwork strip marked as blue in figure 1. In these 3 parts all the mapped features will be focused on from north to south. Other parts of the research area will be discussed by research partners van Gelderen (2021), Melger (2021) and Wesselman (2021).The fourth part of this paragraph will focus on some other results that do not directly fit within the other three parts.
1. Macro
8 As shown in figure … most of the northern part of the fieldwork area has been covered by the sander. The sander mainly consists of medium fine sand which has been deposited like a big alluvial fan by the melting ice masses in and shortly after the Weichselian glaciation period and is therefore mapped as ‘Fluvioglacial accumulative’. On the edge of the sander the village Radeland can be found which has been mapped as ‘Anthropogenic influenced’. When looking at the total research area it becomes clear that humans have mostly built on the parts of the area that contain a lot of sand. This land is less wet than the ice-marginal valley which has a lot of lowland peat soils. The dryer sandy areas like the sander, dunes and moraines are also less suitable for agricultural purposes. Later in time the sander has been undercut by the ice-marginal valley (Juschus, 2001) creating a sharp line between the sander and the ice-marginal valley marked with ‘B’ in figure 5.
Figure 5: A: Radeland, B: Edge of the sander.
Moving south from the sander the ice-marginal valley can be found. This is the lowest and wettest part of the research area and mainly consists of lowland peat soils. The ice-marginal valley has mainly been formed by melting glacial water and is therefore mapped as
‘fluvioglacial erosive’ and ‘fluvioglacial erosive valley’ depending on the terrace level. The south of the research area consists of ground moraines and terminal moraine which have been formed by the ice masses during the saalian glaciation. The terminal moraines were formed right in front of the ice masses and therefore mark the furthest extension of the ice masses during the formation of this area. The ground moraines were deposited from underneath the ice mass and lay in between the ice-marginal valley and terminal moraines. Since the ground and terminal moraines have been deposited by the ice masses they are mapped as ‘glacial accumulative ground moraines’ and ‘glacial accumulative terminal
9 moraines’. Also the ground moraines have been undercut by the ice-marginal valley which has caused sharp distinction between the elevation of both areas.
After the formation of the previously mentioned landforms aeolian processes have formed dune formations on top of the ice-marginal valley, terminal moraines and ground moraines. In the middle of the research strip on the ice-marginal valley there seems to be a longitudinal dune formation. However when looking on a smaller scale some parabolic dunes can be distinguished which have been heavily affected by anthropogenic processes. Further south parallel to the ground moraines a remarkable dune formation can be seen which will be discussed in the fourth part of this paragraph and in the discussion. Even further down a big dune formation, consisting of several large and small parabolic dunes, covers the terminal moraines, ground moraines and ice-marginal valley.
2. Meso
Figure 6: Overview of the Meso features of the research strip.
On a meso level the first thing to be seen in the north of the research area are the small v shaped valleys leading from the sander to the lower ice-marginal valley. A little bit further south on the border between the sander and the ice-marginal valley a main road can be found which has likely been built there for the same reason most buildings are built near dunes, the sander or the moraines. The dry sands are more suited to build on than the wet ice-marginal valley. In the ice-marginal valley a lot of perennial artificial drainageways can be found which are used for the agricultural activities in the ice-marginal valley. Just like on the macro dunes have been mapped on a meso scale. However the meso scale has been
10 focussed on individual dunes instead of aeolian affected areas. Also in the south of the research area strip dunes have been identified, however these show much more clearly the parabolic shapes than the dunes found in the middle of the ice-marginal valley. Furthermore the south contains a wide v-shaped valley leading from the ground moraines to the ice-marginal valley. Also another main road can be found running parallel to the ground moraines for the same reason as previously mentioned.
3. Micro
Figure 7: Overview of the Micro features of the research strip.
The most prevalent features on a micro scale in the north of the research area strip are the roads. Most roads are located on the sander to provide access to some tiny houses on the sander or to accommodate forestry. Another prevalent feature in the north of the research area is the big alluvial fan that runs from the sander onto the ice-marginal valley. The village Radeland is also built on top of this alluvial fan. Within a small valley in the alluvial fan a suspected Charcoal work has been mapped. And a kettle can be found somewhat east of the alluvial fan still on the sander. Further south onto the ice-marginal valley a military
structure has been mapped which, based on the shape and size, may have been used as an artillery station during the second world war. Straight south from this military structure some earth dams/walls have been mapped. Since the earth dams are located in a forest area the purpose is somewhat unclear and requires more research. On micro scale the ridges of the dunes have been mapped which clearly show the different dune types in the research area. Starting with a combination of longitudinal and small parabolic dunes on the ice marginal valley that have been excavated to make way for roads and for the glass production in the
11 village GlasHutte which is located just east of the research strip. Further south parallel to the ground moraines are the somewhat pointy shaped dune formations that will be mentioned later. All the way in the south of the area the parabolic dunes are clearly visible in the ridges of the dunes. In the south of the research strip also more suspected charcoal hearts are mapped.
4. Other Results Paleo rivers
In the south of the research strip a very small paleo river has been found which can be seen in figure 8.
Figure 8: Small paleo river visible on DEM.
However further to the east in the research area a much bigger paleo river has been identified (figure 9). According to de Boer (personal communication, 7-4-2021) this must be the old Dahme river. It can also clearly be seen how originally the river ran north but it has been cut off by the dunes that formed later.
12 Figure 9: Big paleo river as seen on DEM.
13 Spit-like dunes
Parallel to the ground moraines in the south of the ice-marginal valley, dunes can be found which do not match the description of the parabolic or longitudinal dunes like the other dunes in the research area. The crest of the dunes seems to run parallel to the ground moraines and the dunes seem to have some northeast facing teeth like shapes. The appearance of these dunes seems to be somewhat similar to that of spit dunes which are found in coastal delta regions. However these can’t be identified as spit dunes because the ice-marginal valley has never been a coastal region. Possible explanations for the formation of these remarkable dunes can be found in the discussion.
Figure 10: Spit dunes in a deltaic coast. (Ranwell, 1977)
Figure 11: Unidentified dunes in the research area.
14 Mistake in legend 2020
One of the goals of this research was to improve the legend for the geomorphological map that had been created by Schadee (2020). One of the required improvements was the addition of the legend unit ‘fluvioglacial accumulative’ to correctly map the sander in the north east of the research area. Like described in the methods, additional legend items would be based on the legend created by Frank (1987). However, the code for ‘fluvioglacial accumulative’ was ‘13.6’ in the legend of Frank (1987), which had been used by Schadee (2020) for ‘fluvioglacial erosive’. This was most likely done to differentiate between different terrace levels of the ice-marginal valley and didn’t form an issue since no fluvioglacial accumulative landforms were present in the research area of 2020. In order to solve this issue the two terrace levels have been split up into ‘13.7.1’ and ‘13.7.2’ leaving ‘13.6’ available for ‘fluvioglacial accumulative’.
Discussion
No fieldwork possibleDue to Covid-19 related travel restrictions during this research it has not been possible to do fieldwork in the research area. According to Geskus (2020) it is important to check digitally mapped features in the field to increase the reliability of the geomorphological map.
Especially smaller landforms like military structures and relic charcoal works are difficult to identify with a high certainty using sole digital sources. However also the marco features could be mapped more accurately if fieldwork would have been possible since most of the digital data only provides information on the top layer of the surface while a more accurate map would require some knowledge of underlying layers in the soil. Checking
geomorphological features in the field is something that could be done in future research. Mapping Dunes
According to Hugenholtz et al (2012) the mapping of dunes using remote sensing data like LiDAR can be very subjective and offers room for interpretations. To adjust for this the dunes in the research area have been mapped in three different ways on macro, meso and micro level, showing the total aeolian influenced area’s on macro level, the dune shapes on meso level and the ridges of the dunes on micro level. However, like figure … shows within these levels there is still room for different results.
15 Figure 12: Different ways of mapping the same dune. (Hugenholtz et al, 2012)
In order to map the 4 river terraces that Juschus (2001) mentions, multiple cross sections of the ice-marginal valley had been made. However despite the accuracy of the DEM ArcGIS Pro didn’t give an accurate result making it hard to distinguish different terraces. Also a feature that shows a location in the cross section on the map could not be found within ArcGIS Pro. Using Google Earth pro which does include that feature didn’t give an accurate result either because the elevation data didn’t seem accurate and was heavily affected by trees and houses. In the end the best way to visualise and map the different terraces was the method described by van Gelderen (2021).
Spit-like Dunes
The oddly shaped dunes running parallel to the ground moraines which have previously been mentioned could potentially be explained by a few processes which have taken or are still taking place in the research area. As previously mentioned the shape of the dunes shows similarities with that of spit dunes formed in a deltaic coast. Despite this not being a coastal region it might have been a delta which formed from the erosive valleys in the Saalian ground moraines in the south west. This argument might be substantiated by the fact that these dunes are found directly in line with two very big erosive valleys in the moraines. However they do not seem to be present near other valleys in the ground
moraines. Something else that might have played a role in the formation of these dunes are bidirectional winds. The North-South and East-West directions of the arms of the dunes suggest winds from the north and east which could have been possible. During the glacial and early periglacial periods which have shaped most of the landforms in the research area the wind was most likely predominantly coming from the east. According to de Boer
(personal communication, 28-5-2021) this was the case because of the ice masses in the north east causing a high pressure zone in comparison to the warmer south west. When the ice masses were close to the research area the wind direction was most likely predominantly north. But when the ice masses retracted further north the wind direction got predominantly east because of the Coriolis effect. After the glacial periods this wind direction has changed to being predominantly west. Another remarkable feature of these dunes are the steep
16 slopes on the north and east side and the relatively shallow slope on the south and west side. Some dune types like barchan dunes and parabolic dunes naturally have a steeper side facing the same direction as the wind is blowing. In this case that would suggest winds from the south-west. However there could also be several other reasons for the steeper north and east sides. Just like fluvio-glacial erosion from water in the ice-marginal valley undercut the moraines and sander it could, in a later stadium, also have undercut the dunes on the sides facing the ice-marginal valley. Another explanation would be excavation by humans. According to de Boer (personal communication, 28-5-2021) sand was a useful material to stabilize the wet peat soils or make peat soils more fertile in the ice-marginal valley. For this reason most of the anthropogenic structures like villages and cities in the research area are found on or near Aeolian influenced locations.
Valley Terraces
According to Juschus (2001) the ice-marginal valley has 4 different terraces of which the highest and oldest are in the south, and the lower and younger terrasses are in the north. Two of these terraces could easily be identified and have been mapped on the
geomorphological map. In order to map the 4 river terraces that Juschus (2001) mentions, multiple cross sections of the ice-marginal valley had been made. However despite the accuracy of the DEM ArcGIS Pro didn’t give an accurate result making it hard to distinguish different terraces. Also a feature that shows a location in the cross section on the map could not be found within ArcGIS Pro. Using Google Earth pro which does include that feature didn’t give an accurate result either because the elevation data didn’t seem accurate and was heavily affected by trees and houses. Meaning that the 4 different terraces as described by Juschus (2001) could not easily be distinguished with the methods described in this research. The methodology described by research partner van Gelderen might be more effective to make a distinction between the different terraces.
Separating Feature Class
As mentioned in ‘methods and data’ last year's feature class “Unit 13” containing macro geomorphological features has been separated into different feature classes to improve the accuracy by accommodating overlapping features. However this adjustment also came with some challenges. The first of which is accurately mapping geomorphological features that are covered by another feature. For example a dune can lay on top of terminal moraines and ground moraines which makes it hard to map the exact border between the moraine types. This is especially difficult when using digital sources only. This means that this adjustment has made the map better at representing the situation but that it also has created some less accurate spots in the map where multiple features overlap each other. A second challenge that has been caused by this adjustment is that features can no longer be mapped by splitting them from one big polygon. This means that features have to be created from their own separate feature class and snapped to other features to create a fully covered map. This makes the workflow more time consuming and more sensitive to drawing errors.
17
Conclusion
To conclude a geomorphological map has successfully been created showing the geomorphology of the research area on a macro, meso and micro scale. Some dune formations could not be identified as a certain dune type but shows resemblance with spit dunes found in coastal regions. There are also multiple potential explanations for the formations of these dunes. Because traveling to the research area during this research was not possible, not all identified features can be confirmed. This should be done in future research to make the created map more reliable
Literature
Boer, W. M. de (1995). Äolische Prozesse und Landschaftsformen im mittleren Baruther Urstromtal seit dem Hochglazial der Weichselkaltzeit, DOI: 10.18452/13573
Frank, F. (1987). Die Auswertung grossmassstabiger Geomorphologischer Karten (GMK 25) fur den Schulunterricht. Berliner Geographische Abhandlungen, 46(Gmk 25).
https://doi.org/https://doi.org/10.23689/fidgeo-3192
Van Gelderen, H. (2021) the reliability of digital classification and mapping of the Baruth Ice-marginal valley
Geskus, S. (2020) The reliability of geomorphological mapping using LiDAR data: the Central Baruth Ice-Marginal Valley (unpublished
bachelor thesis). University of Amsterdam, Amsterdam, Netherlands. Retrieved from http://www.gisstudio.nl/index.php?page=bsc#geskus
Hugenholtz, C. H., Levin, N., Barchyn, T. E., & Baddock, M. C. (2012). Remote sensing and spatial analysis of aeolian sand dunes: A review and outlook. Earth-Science Reviews, 111(3–4), 319–334. https://doi.org/10.1016/j.earscirev.2011.11.006
Juschus, O. (2001). Das Jungmoränenland südlich von Berlin - Untersuchungen zur jungquartären Landschaftsentwicklung zwischen Unterspreewald und Nuthe.
https://doi.org/10.18452/14585
Jones, A. F., Brewer, P. A., Johnstone, E., & Macklin, M. G. (2007). High-resolution interpretative geomorphological mapping of river valley environments using airborne LiDAR data. Earth Surface Processes and Landforms, 32(10), 1574–1592.
https://doi.org/10.1002/esp.1505
Luimes, B.J. (2020) The adequacy of digital geomorphological research - Research into the adequacy of digital geomorphological conducted research by creation of a
geomorphological map and identification of ridge and furrow-systems, for Horstwalde, Germany. University of Amsterdam, Amsterdam, Netherlands. Retrieved from
18 Melger, A. (2021) The creation of a geomorphology map and the identification of the paleo drainage system of the Hammerfließ stream in the Central Baruth Ice-Marginal Valley with the use of LiDAR Data
Pachur, H. & Schulz, G. (1983). Erläuterungen zur Geomorphologischen Karte 1:25 000 der Bundesrepublik Deutschland GMK 25 Blatt 13, 3545 Berlin-Zehlendorf, P. 1-88.
Romar, M. (2020) Digitally mapping the geomorphology of the Baruth Ice-Marginal Valley, Germany (unpublished bachelor thesis). University of Amsterdam, Amsterdam,
Netherlands. Retrieved from http://www.gis-studio.nl/index.php?page=bsc#romar
Ranwell, D.S. & Boar, R., (1977). Coast dune management guide. Institute of Terrestrial Ecology, HMSO, London.
Schadee, M. (2020). Creating a geomorphological map of a formerly glaciated area in Brandenburg, Germany (unpublished bachelor thesis). University of Amsterdam, Amsterdam, Netherlands. Retrieved from
http://www.gisstudio.nl/index.php?page=bsc#Schadee
Wesselman, J. (2021) Geomorphological mapping and tracing of paleo-river systems in Baruth Ice Marginal Valley, Brandenburg, Germany – By use of LiDAR data, satellite images in ArcGIS Pro and conventional geological data
Zuidervaart, S.J.C (2020) The creation of a large scale Geomorphological map of the Central Baruth Ice-Marginal Valley, Germany (unpublished bachelor thesis). University of
Amsterdam, Amsterdam, Netherlands. Retrieved from http://www.gisstudio.nl/index.php?page=bsc#zuidervaart
Acknowledgements
First I would like to thank dr. W.M. de Boer for providing an incredible amount of data and information about the area, providing feedback, meeting and discussing multiple times during the research and helping to move the research in the right direction. Secondly I would like to thank research partners Jaap Wesselman, Aletta Melger and Hein van Gelderen for the pleasant teamwork during this research.
19
Appendices
21 Data Used for Features
Macro
Fluvioglacial ground
moraine GÜK-100 Teltow flaming
Fluvioglacial terminal
moraine GÜK-100 Teltow flaming
Fluvioglacial erosive GÜK-100 Teltow flaming & Contour Fluvioglacial erosive
valley GÜK-100 Teltow flaming & Contour & DEM
Aeolian Satelite imagery, GÜK-100 Teltow flaming, Aspect, DEM
Anthropogenic
Structures Satelite imagery, DEM
Anthropogenic
Influenced Satelite imagery
Loamy sand
Bodenarten und Substrate – INSPIRE View-Service (WMS-LBGR-BOARTSUBSTR
Lowland Peat
Bodenarten und Substrate – INSPIRE View-Service (WMS-LBGR-BOARTSUBSTR
Sand medium fine
Bodenarten und Substrate – INSPIRE View-Service (WMS-LBGR-BOARTSUBSTR
Sand fine
Bodenarten und Substrate – INSPIRE View-Service (WMS-LBGR-BOARTSUBSTR
Meso Hydrology DEM, Hillshade, Satalite imagery, WMS BB-BE DTK10 Farbe
Dunes DEM, Aspect
Roads WMS BB-BE DTK10 Farbe
Valley & drainage
ways DEM, Aspect
Micro Roads WMS BB-BE DTK10 Farbe
Military DEM, Hillshade, ingescande kaart
Houses https://download.geofabrik.de/europe/germany/brandenburg.html
Dunes DEM, Aspect
