Background
Archaeological sites are increasingly threatened by climate-related hazards. In respon... more Background Archaeological sites are increasingly threatened by climate-related hazards. In response, heritage management authorities initiated projects to document damage and plan risk assessment measures. We present a project initiated after the heavy rainfall and subsequent flood event of July 2021, which involved extensive fieldwork to document the damage to archaeological sites in the Rhineland. We use this database to characterise and assess the damage and investigate site-specific and geospatial factors to identify potential predictive parameters for site damage. Results During fieldwork, we found that the flood damaged 19% of the 538 archaeological sites surveyed. The majority of damaged sites are relatively recent, dating from the medieval or modern periods, and are associated with the use of water power. Damage was mainly caused by erosion, floating debris and washout, e.g. mortar. In a case study, we tested the option of comparing pre- and post-disaster Airborne Laser Scanning elevation data to identify damages. It showed that not only the damage detected during fieldwork was found but also additional areas of loss. In general, however, and quantified based on the entire dataset, the ordnance survey Airborne Laser Scanning data were of limited use for monitoring flood-related damage and could not replace fieldwork. Our statistical analysis of possible risk factors, including both site characteristics and geospatial parameters, using Naïve Bayes Modelling and chi-squared tests, showed that no set of parameters could consistently predict the preservation or damage of archaeological sites across all catchments. In contrast, some external geospatial factors correlated with the occurrence of damage. Conclusions The study highlights both the strengths and limitations of the approaches used to assess and predict the damage to the archaeological heritage in the 2021 flood zones of the Rhineland. It also demonstrates the complexity of the data and spatial processes involved, which limits generalisation but can still inform decision-making for archaeological site management and on-site protection measures in flood-prone areas. With the prospect of more frequent heavy rainfall due to climate change, the specific needs of the archaeological heritage should be integrated into broader prevention and disaster management plans.
Since the first years, elevation data has played an important role in many contributions presente... more Since the first years, elevation data has played an important role in many contributions presented at the conferences in Vienna. The 25 th conference which formerly had the German title "Archäologie und Computer" and now is known as "Cultural Heritage and New Technologies" is the occasion for looking back on early archaeological applications of digital elevation models (DEMs), advances in the course of time, the state of the art and future developments. Many of the approaches presented rely on elevation data grids that are generated from irregularly distributed altitude points. Before the first conference in Vienna, several high-impact papers dealing with elevation data were published. In most cases, their goal was to create predictive models, i.e., to delimit areas of high probability for detecting an archaeological site. Attributes derived from elevation data such as slope, aspect, view quality and shelter were often used as predictor variables. The papers presented at the Vienna conferences are evidence of the fact that archaeological predictive modelling is a highly controversial topic. Additional contributions at the conference in Vienna are revisited that deal with retrodictive modelling, i.e., explaining the distribution of known sites based on landscape properties, with predicting (or retrodicting) the location of linear features such as roads or boundaries, and with issues of low-resolution digital elevation models. Moreover, contributions at the Vienna conference are highlighted that reflect the development of applying high resolution elevation data for detecting archaeological sites. The examples presented are supplemented by some illustrations from the author's and her colleague's work in the Rhineland.
An important and frequently used tool in archaeological spatial analysis is least-cost path (LCP)... more An important and frequently used tool in archaeological spatial analysis is least-cost path (LCP) analysis with the aim of computing routes connecting a set of targets. The outcome depends on the cost model chosen and the topographic data used. A slope-dependent cost model requires a digital elevation model (DEM) that should reflect the landscape in the past. Often, it is not possible to reconstruct the past terrain, and modern high resolution elevation data results in problematic storage requirements and computation times. This article presents a case study that explores issues in replication and stability of LCP calculations for pairs of targets that are close to known old trade routes. These trade routes are used for evaluating a large number of cost models, that are based on two topographic data sets including DEMs of two different resolutions. Six different slope-dependent cost functions for pedestrians are tested as well as a function modeling wheeled transport. The latter generates hairpin curves when steep gradients exceeding a predefined critical value are to be mastered. By introducing additional isotropic cost factors, the initial slope-dependent cost model is refined iteratively in terms of the performance indicators presented. Surprisingly, overall improvement in the performance indicators often leads to worse replication of some of the known trade routes. The overall best model found is optimal for only eight out of 19 trade route sections considered. This is probably due to different functions of the routes. The performance of the best cost model based on the DEM with a cell size of 25 m is not significantly higher than that based on a larger cell size of 50 m. It is shown that the best-performing cost model parameters for the two topographic data sets are closely related but not identical.
M. Gillings, P. Hacigüzeller, G. Lock (eds.) Archaeological Spatial Analysis: A Methodological Guide to GIS. Routledge, Taylor & Francis Group: London and New York, 333-358., 2020
Each model of past movement based on the historical and archaeological evidence nowadays relies i... more Each model of past movement based on the historical and archaeological evidence nowadays relies implicitly or explicitly on a cost function estimating costs of movement in terms of time, calories or some other currency for the study area and period of time considered. Cost functions are the basis of two popular GIS-based areas of spatial analysis in archaeology: site catchments and least-cost paths. A site catchment is the region accessible from a site, and often archaeological studies analyse the resources within this region. A least-cost path (LCP) is the route minimizing the costs of movement between two given locations. In archaeological studies, the aim of LCP calculations is often to reconstruct ancient routes or a route network or to identify the principal factors governing the construction of known roads or road segments. For generating a site catchment, GIS programs compute an accumulated cost surface (ACS). This is also the first step for LCP generation, therefore the Dijkstra algorithm implemented in most GIS programs for this purpose is outlined. The cost function applied for calculating the ACS typically depends on slope, soil type, vegetation cover or the presence of streams. Several slope-dependent cost functions and a list of terrain factors are presented. Moreover, issues with least-cost applications are discussed, including barrier implementation, restricted number of possible movement directions, and the calculation of slope-based costs for a path taken in one or two directions. The aim of the case study is to compare the site catchments of a Roman farm and a temple at different scales based on a cost model derived from Roman roads that were recorded in three publications. The concluding section discusses validation and stability of the outcomes of the methods presented and some additional issues such as modelling movement beyond walking. Finally, some other cost function based approaches are outlined.
Museen der Stadt Wien, Stadtarchäologie (Hrsg.): Monumental Computations: Digital archaeology of large urban and underground infrastructures, Heidelberg: Propylaeum, 2021 (Proceedings of the International Conference on Cultural Heritage and New Technologies, Vienna, Band 24), 2021
National or regional archaeological databases have a long tradition in digital cultural heritage ... more National or regional archaeological databases have a long tradition in digital cultural heritage management. This paper introduces the database integrating a GIS component used in the Rhineland Commission for Archaeological Sites and Monuments, Germany. This data base is an indispensable tool for archaeologists in the Rhineland when monitoring planning projects that cover an area overlapping with archaeological sites.
Museen der Stadt Wien, Stadtarchäologie (Hrsg.): Monumental Computations: Digital archaeology of large urban and underground infrastructures, Heidelberg: Propylaeum, 2021 (Proceedings of the International Conference on Cultural Heritage and New Technologies, Vienna, Band 24), 2021
In May 2018, Lidar data was acquired in a densely vegetated test area in southern Japan with the ... more In May 2018, Lidar data was acquired in a densely vegetated test area in southern Japan with the aim of detecting Sue kiln sites, that date back in the 9 th century. In the region including the test area, several surveys as well as an excavation recorded evidence of such sites in recent years. As the densely forested, mountainous topography complicates any ground based archaeological investigations, airborne Lidar was considered the method of choice for surveying the area. This is the first study of this kind in Japan. The company that acquired the Lidar data also provided lists of ground points. Initial interpolations and visualisations relied on these lists; later the program SCOP++ was applied for generating digital terrain models based on alternative ground point selection strategies. The results of several approaches for visualising and analysing the Lidar data sets will be presented and discussed. For visualisations, mainly low-cost or free software was applied, including GIS software, the Relief Visualisation Toolbox (RVT), and planlauf/TERRAIN for 3D virtual flights. Additional GIS approaches for analysing the data are presented: (1) contour maps that assist navigation in the field, (2) a density map of ground points allows assessing the reliability of the Lidar visualisations, (3) cross sections are useful for validating the features recognized and measuring their depth or height, and (4) slope maps support delimiting manmade terraces, identifying platforms and are an important input for deductive predictive modelling of Sue kiln sites. The work of mapping probable kiln locations detected in the LiDAR data and verifying these sites by traditional prospection methods is still in progress. Only after reliably identifying a large number of Sue kilns in the test area, approaches such as predictive modelling or machine learning may be applied for locating additional kilns in this region.
The initial aim of this project was to test, if bou ndaries of the Roman administration units in ... more The initial aim of this project was to test, if bou ndaries of the Roman administration units in a larg e part of southern Germany could be approached mathematically based on kn wn central locations and least-cost Thiessen po lygon calculations. The boundaries derived from SRTM elev ation data by applying the popular Tobler cost func tion roughly coincide with those of standard Thiessen po lygons even in hilly regions. These results do not change significantly when choosing a slope-dependent cost function for vehicles rather than for pedestrians. For this reason, we decided to investigate the issue more thoroughly in a small part of the study area. For this area, we mapped some traditional road reconstructions of Roman roads. Th e known Roman roads allow estimating the costs of m ove ent during Roman times in this part of Germany, i.e. fo r several cost models, least-cost paths (LCP) were compared with the routes of the known Roman roads. Based on the b est performing cost model, ac...
Proceedings of the 23rd International Conference on Cultural Heritage and New Technologies 2018, 2019
According to a historical document dating back to 1065 AD, the medieval road known as strata Colo... more According to a historical document dating back to 1065 AD, the medieval road known as strata Coloniensis connected the monastery of Essen-Werden with Cologne in Germany. Several alternative hypotheses concerning the course of this route were published, they only agree on a fairly small section south of Essen-Werden. The aim of the paper is to identify the most probable course of the medieval road between Essen-Werden and the river Düssel. In their arguments, the supporters of two hypotheses refer to historical maps. After outlining the hypotheses, the paper discusses these historical maps and additional earlier maps, several of them are available in online repositories. For each of the early maps considered, the roads or paths depicted were digitized, mostly on the basis of place names and water courses that could be found both on the old and a more recent map. The results allow assessing the accuracy of the maps and investigating the continuity of the roads. Visualization of high resolution elevation data and aerial images were applied to identify old road sections in some areas. However, these approaches are limited by substantial modern landscape modifications and a large proportion of built-up areas. Due to major changes of the relief, reliable least-cost path calculations to reconstruct the old route should not be based on modern digital elevation data. But it is still possible to estimate the effort of using an old road section by cost computations based on modern topographic data. A comparison of the different route alternatives is presented using these estimations.
Proceedings of the 22nd International Conference on Cultural Heritage and New Technologies 2017. CHNT 22, 2017 (Vienna 2019). http://www.chnt.at/proceedings-chnt-22/, 2019
For the island Curaçao in the southern Caribbean Sea, the defense infrastructure played an import... more For the island Curaçao in the southern Caribbean Sea, the defense infrastructure played an important role in the 17th and even more so in the 18th century. After the Dutch had conquered the island in 1634, they immediately started constructing the first fortification. According to our research based on the book by Hartog (1997), historical reports and a large set of historical maps, 34 fortification sites were established until 1801, some of these lasted only a few days. Several site locations were extended, rebuilt, or moved in the course of time. Most of these sites were located close to bays and natural harbors that allowed landing of smugglers or enemy boats. The remnants of some of these forts can be visited today, but a large proportion of the forts disappeared from the modern maps. Based on the sources mentioned above we were able to narrow down the location of 21 out of 27 disappeared fortresses with an estimated accuracy of 200 m or less. The paper also investigates location factors for the reconstructed fortification locations, i.e. the viewsheds within the cannon ranges. These suggest that some of the fortification positions initially chosen should be moved in order to generate larger viewsheds. The tentative results presented could be improved if high resolution aerial photographs, altitude and topographical data become available. https://www.chnt.at/wp-content/uploads/eBook_CHNT22_Herzog_Mom.pdf
Mitteilungen des Deutschen Archäologen-Verbandes e.V. 49(2), 2018
Einleitung: Das LVR-Amt für Bodendenkmalpflege im Rheinland ist in seinem Arbeitsgebiet zuständig... more Einleitung: Das LVR-Amt für Bodendenkmalpflege im Rheinland ist in seinem Arbeitsgebiet zuständig für den Schutz, die Erfassung und Erforschung des reichen kulturellen Erbes sowie der Zeugnisse tierischen und pflanzlichen Lebens aus erdgeschichtlicher Zeit. Das Arbeits-gebiet umfasst im Land Nordrhein-Westfalen die Regierungsbezirke Köln und Düsseldorf, nicht aber die Stadt Köln (Abb. 1). Der Hauptsitz des Amtes ist in Bonn, hinzu kommen vier Außenstellen. Das Fachamt arbeitet eng mit dem LVR-LandesMuseum in Bonn zusammen, in dem die Funde aus dem Rheinland aufbewahrt und zum Teil ausgestellt werden. Bei der Bestimmung von Münzen und bei Restaurierungsbedarf greift das Fachamt auf Dienstleistungen des LandesMuseums zurück. Beide Institutionen haben eine gemeinsame Geschichte, die Trennung erfolgte erst zu Beginn des Jahres 1987.
Proceedings of the 22nd International Conference on Cultural Heritage and New Technologies 2017, 2019
The initial aim of this project was to test, if boundaries of the Roman administration units in a... more The initial aim of this project was to test, if boundaries of the Roman administration units in a large part of southern Germany could be approached mathematically based on known central locations and least-cost Thiessen polygon calculations. The boundaries derived from SRTM elevation data by applying the popular Tobler cost function roughly coincide with those of standard Thiessen polygons even in hilly regions. These results do not change significantly when choosing a slope-dependent cost function for vehicles rather than for pedestrians. For this reason, we decided to investigate the issue more thoroughly in a small part of the study area. For this area, we mapped some traditional road reconstructions of Roman roads. The known Roman roads allow estimating the costs of movement during Roman times in this part of Germany, i.e. for several cost models, least-cost paths (LCP) were compared with the routes of the known Roman roads. Based on the best performing cost model, accessibility maps were created. Linear features of low accessibility in these maps indicate possible locations of past or present boundaries because the boundaries are often aligned with topographical elements such as steep slopes or rivers. Some issues detected in the course of this research will be discussed, e.g. the difficulties of attributing appropriate costs for crossing the rivers and creeks based on the available data on modern water bodies and the drawbacks of least-cost Thiessen polygon approaches.
Viele Bodendenkmäler wie Landwehren, Grabhügel oder Hohlwege zeichnen sich oberirdisch im Gelände... more Viele Bodendenkmäler wie Landwehren, Grabhügel oder Hohlwege zeichnen sich oberirdisch im Geländerelief ab. In den letzten Jahren bilden zunehmend flugzeugbasierte Laserscanning-Daten (Light detection and ranging, kurz LiDAR) der Landesvermessung die Grundlage für die Entdeckung und Abgrenzung solcher Bodendenkmäler. LiDAR-Daten bestehen aus unregelmäßig verteilten Höhenpunkten auf der Erdoberfläche. Damit erlauben sie im Gegensatz zu Luftbildern auch in Waldgebieten einen von Bäumen ungestörten Blick auf die dort befindlichen Bodendenkmäler. Diese sind durch die Bewaldung teilweise noch obertägig erhalten, aber keineswegs geschützt: Rückegassen der Waldwirtschaft, Mountainbike-Routen, aber auch tief wurzelnde Bäume gefährden ihren Erhalt. Deshalb ist die Auswertung von LiDAR-Daten in Waldgebieten für das LVR-Amt für Bodendenkmalpflege im Rheinland besonders wichtig.
Noch in der frühen Neuzeit war die Landwirtschaft im Bergischen Land mit Abstand der wichtigste W... more Noch in der frühen Neuzeit war die Landwirtschaft im Bergischen Land mit Abstand der wichtigste Wirtschaftsfaktor. Bei der Analyse des Siedlungsmusters auf Grundlage von historischen Karten aus dem Jahr 1715 stellte sich die Frage, wie groß die Wirtschaftsfläche eines Einzelhofs bzw. einer bäuerlichen Siedlung in dieser Zeit war. Nach gängiger Lehrmeinung besteht eine landwirtschaftliche Nutzfläche aus mehreren Zonen, die ringartig um die Siedlungsstelle angeordnet sind, angefangen mit den Hofstellen selbst und den zugehörigen Hausgärten über einen Obstbaumgürtel, Wiesen und Weiden, Ackerland bis hin zu einer Randzone, häufig mit Waldnutzung. Bei der Suche nach Angaben zur Größe der Zonen in der Fachliteratur ließen sich nur Zahlen zur Nutzfläche von mittelalterlichen Hofstellen im Norden Deutschlands finden; diese liegen zwischen 2,5 und 8 ha. Dabei ist nicht klar, wie viele der Zonen Berücksichtigung finden.
Materialien zur Bodendenkmalpflege im Rheinland. LVR-Amt für Bodendenkmalpflege im Rheinland Das Ganze ist mehr als die Summe seiner Teile Das Ganze ist mehr als die Summe seiner Teile. Festschrift für Jürgen Kunow, 2018
Materialien zur Bodendenkmalpflege im Rheinland LVR-Amt für Bodendenkmalpflege im Rheinland Das Ganze ist mehr als die Summe seiner Teile Das Ganze ist mehr als die Summe seiner Teile. Festschrift für Jürgen Kunow, 2018
The aim of this contribution is on the one hand to map pre-industrial long distance roads located... more The aim of this contribution is on the one hand to map pre-industrial long distance roads located in a hilly region east of Cologne, Germany as exactly as possible and on the other hand to assess the accuracy of least-cost approaches that are increasingly applied by archaeologists for prehistoric road reconstruction. Probably the earliest map covering the study area east of Cologne dates back to 1575. The map is distorted so that rectification is difficult. But it is possible to assess the local accuracy of the map and to transfer the approximate routes to a modern map manually. Most of the area covered by the 1575 map is also depicted on a set of more accurate maps created in the early 19th century and a somewhat later historical map set (ca. 1842 AD). The historical roads on these rectified historical maps close to the approximate roads were digitized and compared to the outcomes of least-cost analysis, specifically least-cost paths and accessibility maps. Based on these route reconstructions with limited accuracy, Lidar data is checked to identify remains of these roads. Several approaches for visualizing Lidar data are tested to identify appropriate methods for detecting sunken roads. Possible sunken roads detected on the Lidar images were validated by checking cross sections in the digital elevation model and in the field.
For part of Bergisches Land east of Cologne, Germany, the maps by Ploennies finished in 1715 AD a... more For part of Bergisches Land east of Cologne, Germany, the maps by Ploennies finished in 1715 AD and the accompanying text provide detailed data for testing approaches used in archaeology for determining the boundaries of agrarian production. The historical maps indicate the settlement sizes ranging from single farmsteads to towns with walls. In 1715 AD the towns were quite small and the economy in this rural area relied mostly on agriculture (oats and livestock). The boundaries of agrarian production can be estimated by applying methods for allocating farmland territories to the settlement locations. The methods discussed are site catchments, retrodictive modelling for neighbourhoods, Thiessen polygons and approaches based on the distances between neighbouring settlement locations. With the exception of retrodictive modelling for neighbourhoods, two variants of the approaches are applicable, either resting on straight-line or least-cost distances.
Digital elevation models (DEMs) are an important basis for many archaeological GIS studies like p... more Digital elevation models (DEMs) are an important basis for many archaeological GIS studies like predictive modelling, visibility and least-cost path analysis. Issues in DEM data have been discussed in textbooks dealing with archaeological GIS applications and in some papers published more than ten years ago. But lately hardly any archaeological GIS study relying on DEM data refer to these issues though the problems connected with DEM use have not decreased with increasing availability of DEM data. This contribution presents a case study in the East Andean slopes of Ecuador analysing the impact of DEM choice on slope and aspect calculation as well as least-cost site catchments and least-cost paths. Four freely available elevation data sets are considered: ASTER GDEM2, SRTM at 3 arc-second and 1 arc-second resolution as well as a DEM derived from digitised contour lines of the official 1:50,000 maps provided by the Military Geographical Institute of Ecuador. Issues discussed are DEM resolution, horizontal and vertical accuracy, filling voids, and creating a DEM from contour lines.
CAA 2015 - Keep the Revolution Going. Proceedings of the 43rd Annual Conference on Computer Applications and Quantitative Methods in Archaeology held in Siena 2014
People who walk into unknown territory see only part of the landscape ahead whereas optimal path ... more People who walk into unknown territory see only part of the landscape ahead whereas optimal path calculation relies on the assumption that the walker has total knowledge of the landscape ahead. For this reason optimal path calculations are not appropriate for modelling dispersal processes. But the cost function derived from optimal path calculations can also be applied for modelling dispersal: The walker will prefer paths consisting of locally optimal steps in the direction chosen initially. Based on this assumption, a new agent-based algorithm is proposed for modelling the dispersal into unknown terrain starting from a given location. The corridors of movement generated this way are compared to globally optimal radial paths for a hilly study area located east of Cologne in Germany.
Background
Archaeological sites are increasingly threatened by climate-related hazards. In respon... more Background Archaeological sites are increasingly threatened by climate-related hazards. In response, heritage management authorities initiated projects to document damage and plan risk assessment measures. We present a project initiated after the heavy rainfall and subsequent flood event of July 2021, which involved extensive fieldwork to document the damage to archaeological sites in the Rhineland. We use this database to characterise and assess the damage and investigate site-specific and geospatial factors to identify potential predictive parameters for site damage. Results During fieldwork, we found that the flood damaged 19% of the 538 archaeological sites surveyed. The majority of damaged sites are relatively recent, dating from the medieval or modern periods, and are associated with the use of water power. Damage was mainly caused by erosion, floating debris and washout, e.g. mortar. In a case study, we tested the option of comparing pre- and post-disaster Airborne Laser Scanning elevation data to identify damages. It showed that not only the damage detected during fieldwork was found but also additional areas of loss. In general, however, and quantified based on the entire dataset, the ordnance survey Airborne Laser Scanning data were of limited use for monitoring flood-related damage and could not replace fieldwork. Our statistical analysis of possible risk factors, including both site characteristics and geospatial parameters, using Naïve Bayes Modelling and chi-squared tests, showed that no set of parameters could consistently predict the preservation or damage of archaeological sites across all catchments. In contrast, some external geospatial factors correlated with the occurrence of damage. Conclusions The study highlights both the strengths and limitations of the approaches used to assess and predict the damage to the archaeological heritage in the 2021 flood zones of the Rhineland. It also demonstrates the complexity of the data and spatial processes involved, which limits generalisation but can still inform decision-making for archaeological site management and on-site protection measures in flood-prone areas. With the prospect of more frequent heavy rainfall due to climate change, the specific needs of the archaeological heritage should be integrated into broader prevention and disaster management plans.
Since the first years, elevation data has played an important role in many contributions presente... more Since the first years, elevation data has played an important role in many contributions presented at the conferences in Vienna. The 25 th conference which formerly had the German title "Archäologie und Computer" and now is known as "Cultural Heritage and New Technologies" is the occasion for looking back on early archaeological applications of digital elevation models (DEMs), advances in the course of time, the state of the art and future developments. Many of the approaches presented rely on elevation data grids that are generated from irregularly distributed altitude points. Before the first conference in Vienna, several high-impact papers dealing with elevation data were published. In most cases, their goal was to create predictive models, i.e., to delimit areas of high probability for detecting an archaeological site. Attributes derived from elevation data such as slope, aspect, view quality and shelter were often used as predictor variables. The papers presented at the Vienna conferences are evidence of the fact that archaeological predictive modelling is a highly controversial topic. Additional contributions at the conference in Vienna are revisited that deal with retrodictive modelling, i.e., explaining the distribution of known sites based on landscape properties, with predicting (or retrodicting) the location of linear features such as roads or boundaries, and with issues of low-resolution digital elevation models. Moreover, contributions at the Vienna conference are highlighted that reflect the development of applying high resolution elevation data for detecting archaeological sites. The examples presented are supplemented by some illustrations from the author's and her colleague's work in the Rhineland.
An important and frequently used tool in archaeological spatial analysis is least-cost path (LCP)... more An important and frequently used tool in archaeological spatial analysis is least-cost path (LCP) analysis with the aim of computing routes connecting a set of targets. The outcome depends on the cost model chosen and the topographic data used. A slope-dependent cost model requires a digital elevation model (DEM) that should reflect the landscape in the past. Often, it is not possible to reconstruct the past terrain, and modern high resolution elevation data results in problematic storage requirements and computation times. This article presents a case study that explores issues in replication and stability of LCP calculations for pairs of targets that are close to known old trade routes. These trade routes are used for evaluating a large number of cost models, that are based on two topographic data sets including DEMs of two different resolutions. Six different slope-dependent cost functions for pedestrians are tested as well as a function modeling wheeled transport. The latter generates hairpin curves when steep gradients exceeding a predefined critical value are to be mastered. By introducing additional isotropic cost factors, the initial slope-dependent cost model is refined iteratively in terms of the performance indicators presented. Surprisingly, overall improvement in the performance indicators often leads to worse replication of some of the known trade routes. The overall best model found is optimal for only eight out of 19 trade route sections considered. This is probably due to different functions of the routes. The performance of the best cost model based on the DEM with a cell size of 25 m is not significantly higher than that based on a larger cell size of 50 m. It is shown that the best-performing cost model parameters for the two topographic data sets are closely related but not identical.
M. Gillings, P. Hacigüzeller, G. Lock (eds.) Archaeological Spatial Analysis: A Methodological Guide to GIS. Routledge, Taylor & Francis Group: London and New York, 333-358., 2020
Each model of past movement based on the historical and archaeological evidence nowadays relies i... more Each model of past movement based on the historical and archaeological evidence nowadays relies implicitly or explicitly on a cost function estimating costs of movement in terms of time, calories or some other currency for the study area and period of time considered. Cost functions are the basis of two popular GIS-based areas of spatial analysis in archaeology: site catchments and least-cost paths. A site catchment is the region accessible from a site, and often archaeological studies analyse the resources within this region. A least-cost path (LCP) is the route minimizing the costs of movement between two given locations. In archaeological studies, the aim of LCP calculations is often to reconstruct ancient routes or a route network or to identify the principal factors governing the construction of known roads or road segments. For generating a site catchment, GIS programs compute an accumulated cost surface (ACS). This is also the first step for LCP generation, therefore the Dijkstra algorithm implemented in most GIS programs for this purpose is outlined. The cost function applied for calculating the ACS typically depends on slope, soil type, vegetation cover or the presence of streams. Several slope-dependent cost functions and a list of terrain factors are presented. Moreover, issues with least-cost applications are discussed, including barrier implementation, restricted number of possible movement directions, and the calculation of slope-based costs for a path taken in one or two directions. The aim of the case study is to compare the site catchments of a Roman farm and a temple at different scales based on a cost model derived from Roman roads that were recorded in three publications. The concluding section discusses validation and stability of the outcomes of the methods presented and some additional issues such as modelling movement beyond walking. Finally, some other cost function based approaches are outlined.
Museen der Stadt Wien, Stadtarchäologie (Hrsg.): Monumental Computations: Digital archaeology of large urban and underground infrastructures, Heidelberg: Propylaeum, 2021 (Proceedings of the International Conference on Cultural Heritage and New Technologies, Vienna, Band 24), 2021
National or regional archaeological databases have a long tradition in digital cultural heritage ... more National or regional archaeological databases have a long tradition in digital cultural heritage management. This paper introduces the database integrating a GIS component used in the Rhineland Commission for Archaeological Sites and Monuments, Germany. This data base is an indispensable tool for archaeologists in the Rhineland when monitoring planning projects that cover an area overlapping with archaeological sites.
Museen der Stadt Wien, Stadtarchäologie (Hrsg.): Monumental Computations: Digital archaeology of large urban and underground infrastructures, Heidelberg: Propylaeum, 2021 (Proceedings of the International Conference on Cultural Heritage and New Technologies, Vienna, Band 24), 2021
In May 2018, Lidar data was acquired in a densely vegetated test area in southern Japan with the ... more In May 2018, Lidar data was acquired in a densely vegetated test area in southern Japan with the aim of detecting Sue kiln sites, that date back in the 9 th century. In the region including the test area, several surveys as well as an excavation recorded evidence of such sites in recent years. As the densely forested, mountainous topography complicates any ground based archaeological investigations, airborne Lidar was considered the method of choice for surveying the area. This is the first study of this kind in Japan. The company that acquired the Lidar data also provided lists of ground points. Initial interpolations and visualisations relied on these lists; later the program SCOP++ was applied for generating digital terrain models based on alternative ground point selection strategies. The results of several approaches for visualising and analysing the Lidar data sets will be presented and discussed. For visualisations, mainly low-cost or free software was applied, including GIS software, the Relief Visualisation Toolbox (RVT), and planlauf/TERRAIN for 3D virtual flights. Additional GIS approaches for analysing the data are presented: (1) contour maps that assist navigation in the field, (2) a density map of ground points allows assessing the reliability of the Lidar visualisations, (3) cross sections are useful for validating the features recognized and measuring their depth or height, and (4) slope maps support delimiting manmade terraces, identifying platforms and are an important input for deductive predictive modelling of Sue kiln sites. The work of mapping probable kiln locations detected in the LiDAR data and verifying these sites by traditional prospection methods is still in progress. Only after reliably identifying a large number of Sue kilns in the test area, approaches such as predictive modelling or machine learning may be applied for locating additional kilns in this region.
The initial aim of this project was to test, if bou ndaries of the Roman administration units in ... more The initial aim of this project was to test, if bou ndaries of the Roman administration units in a larg e part of southern Germany could be approached mathematically based on kn wn central locations and least-cost Thiessen po lygon calculations. The boundaries derived from SRTM elev ation data by applying the popular Tobler cost func tion roughly coincide with those of standard Thiessen po lygons even in hilly regions. These results do not change significantly when choosing a slope-dependent cost function for vehicles rather than for pedestrians. For this reason, we decided to investigate the issue more thoroughly in a small part of the study area. For this area, we mapped some traditional road reconstructions of Roman roads. Th e known Roman roads allow estimating the costs of m ove ent during Roman times in this part of Germany, i.e. fo r several cost models, least-cost paths (LCP) were compared with the routes of the known Roman roads. Based on the b est performing cost model, ac...
Proceedings of the 23rd International Conference on Cultural Heritage and New Technologies 2018, 2019
According to a historical document dating back to 1065 AD, the medieval road known as strata Colo... more According to a historical document dating back to 1065 AD, the medieval road known as strata Coloniensis connected the monastery of Essen-Werden with Cologne in Germany. Several alternative hypotheses concerning the course of this route were published, they only agree on a fairly small section south of Essen-Werden. The aim of the paper is to identify the most probable course of the medieval road between Essen-Werden and the river Düssel. In their arguments, the supporters of two hypotheses refer to historical maps. After outlining the hypotheses, the paper discusses these historical maps and additional earlier maps, several of them are available in online repositories. For each of the early maps considered, the roads or paths depicted were digitized, mostly on the basis of place names and water courses that could be found both on the old and a more recent map. The results allow assessing the accuracy of the maps and investigating the continuity of the roads. Visualization of high resolution elevation data and aerial images were applied to identify old road sections in some areas. However, these approaches are limited by substantial modern landscape modifications and a large proportion of built-up areas. Due to major changes of the relief, reliable least-cost path calculations to reconstruct the old route should not be based on modern digital elevation data. But it is still possible to estimate the effort of using an old road section by cost computations based on modern topographic data. A comparison of the different route alternatives is presented using these estimations.
Proceedings of the 22nd International Conference on Cultural Heritage and New Technologies 2017. CHNT 22, 2017 (Vienna 2019). http://www.chnt.at/proceedings-chnt-22/, 2019
For the island Curaçao in the southern Caribbean Sea, the defense infrastructure played an import... more For the island Curaçao in the southern Caribbean Sea, the defense infrastructure played an important role in the 17th and even more so in the 18th century. After the Dutch had conquered the island in 1634, they immediately started constructing the first fortification. According to our research based on the book by Hartog (1997), historical reports and a large set of historical maps, 34 fortification sites were established until 1801, some of these lasted only a few days. Several site locations were extended, rebuilt, or moved in the course of time. Most of these sites were located close to bays and natural harbors that allowed landing of smugglers or enemy boats. The remnants of some of these forts can be visited today, but a large proportion of the forts disappeared from the modern maps. Based on the sources mentioned above we were able to narrow down the location of 21 out of 27 disappeared fortresses with an estimated accuracy of 200 m or less. The paper also investigates location factors for the reconstructed fortification locations, i.e. the viewsheds within the cannon ranges. These suggest that some of the fortification positions initially chosen should be moved in order to generate larger viewsheds. The tentative results presented could be improved if high resolution aerial photographs, altitude and topographical data become available. https://www.chnt.at/wp-content/uploads/eBook_CHNT22_Herzog_Mom.pdf
Mitteilungen des Deutschen Archäologen-Verbandes e.V. 49(2), 2018
Einleitung: Das LVR-Amt für Bodendenkmalpflege im Rheinland ist in seinem Arbeitsgebiet zuständig... more Einleitung: Das LVR-Amt für Bodendenkmalpflege im Rheinland ist in seinem Arbeitsgebiet zuständig für den Schutz, die Erfassung und Erforschung des reichen kulturellen Erbes sowie der Zeugnisse tierischen und pflanzlichen Lebens aus erdgeschichtlicher Zeit. Das Arbeits-gebiet umfasst im Land Nordrhein-Westfalen die Regierungsbezirke Köln und Düsseldorf, nicht aber die Stadt Köln (Abb. 1). Der Hauptsitz des Amtes ist in Bonn, hinzu kommen vier Außenstellen. Das Fachamt arbeitet eng mit dem LVR-LandesMuseum in Bonn zusammen, in dem die Funde aus dem Rheinland aufbewahrt und zum Teil ausgestellt werden. Bei der Bestimmung von Münzen und bei Restaurierungsbedarf greift das Fachamt auf Dienstleistungen des LandesMuseums zurück. Beide Institutionen haben eine gemeinsame Geschichte, die Trennung erfolgte erst zu Beginn des Jahres 1987.
Proceedings of the 22nd International Conference on Cultural Heritage and New Technologies 2017, 2019
The initial aim of this project was to test, if boundaries of the Roman administration units in a... more The initial aim of this project was to test, if boundaries of the Roman administration units in a large part of southern Germany could be approached mathematically based on known central locations and least-cost Thiessen polygon calculations. The boundaries derived from SRTM elevation data by applying the popular Tobler cost function roughly coincide with those of standard Thiessen polygons even in hilly regions. These results do not change significantly when choosing a slope-dependent cost function for vehicles rather than for pedestrians. For this reason, we decided to investigate the issue more thoroughly in a small part of the study area. For this area, we mapped some traditional road reconstructions of Roman roads. The known Roman roads allow estimating the costs of movement during Roman times in this part of Germany, i.e. for several cost models, least-cost paths (LCP) were compared with the routes of the known Roman roads. Based on the best performing cost model, accessibility maps were created. Linear features of low accessibility in these maps indicate possible locations of past or present boundaries because the boundaries are often aligned with topographical elements such as steep slopes or rivers. Some issues detected in the course of this research will be discussed, e.g. the difficulties of attributing appropriate costs for crossing the rivers and creeks based on the available data on modern water bodies and the drawbacks of least-cost Thiessen polygon approaches.
Viele Bodendenkmäler wie Landwehren, Grabhügel oder Hohlwege zeichnen sich oberirdisch im Gelände... more Viele Bodendenkmäler wie Landwehren, Grabhügel oder Hohlwege zeichnen sich oberirdisch im Geländerelief ab. In den letzten Jahren bilden zunehmend flugzeugbasierte Laserscanning-Daten (Light detection and ranging, kurz LiDAR) der Landesvermessung die Grundlage für die Entdeckung und Abgrenzung solcher Bodendenkmäler. LiDAR-Daten bestehen aus unregelmäßig verteilten Höhenpunkten auf der Erdoberfläche. Damit erlauben sie im Gegensatz zu Luftbildern auch in Waldgebieten einen von Bäumen ungestörten Blick auf die dort befindlichen Bodendenkmäler. Diese sind durch die Bewaldung teilweise noch obertägig erhalten, aber keineswegs geschützt: Rückegassen der Waldwirtschaft, Mountainbike-Routen, aber auch tief wurzelnde Bäume gefährden ihren Erhalt. Deshalb ist die Auswertung von LiDAR-Daten in Waldgebieten für das LVR-Amt für Bodendenkmalpflege im Rheinland besonders wichtig.
Noch in der frühen Neuzeit war die Landwirtschaft im Bergischen Land mit Abstand der wichtigste W... more Noch in der frühen Neuzeit war die Landwirtschaft im Bergischen Land mit Abstand der wichtigste Wirtschaftsfaktor. Bei der Analyse des Siedlungsmusters auf Grundlage von historischen Karten aus dem Jahr 1715 stellte sich die Frage, wie groß die Wirtschaftsfläche eines Einzelhofs bzw. einer bäuerlichen Siedlung in dieser Zeit war. Nach gängiger Lehrmeinung besteht eine landwirtschaftliche Nutzfläche aus mehreren Zonen, die ringartig um die Siedlungsstelle angeordnet sind, angefangen mit den Hofstellen selbst und den zugehörigen Hausgärten über einen Obstbaumgürtel, Wiesen und Weiden, Ackerland bis hin zu einer Randzone, häufig mit Waldnutzung. Bei der Suche nach Angaben zur Größe der Zonen in der Fachliteratur ließen sich nur Zahlen zur Nutzfläche von mittelalterlichen Hofstellen im Norden Deutschlands finden; diese liegen zwischen 2,5 und 8 ha. Dabei ist nicht klar, wie viele der Zonen Berücksichtigung finden.
Materialien zur Bodendenkmalpflege im Rheinland. LVR-Amt für Bodendenkmalpflege im Rheinland Das Ganze ist mehr als die Summe seiner Teile Das Ganze ist mehr als die Summe seiner Teile. Festschrift für Jürgen Kunow, 2018
Materialien zur Bodendenkmalpflege im Rheinland LVR-Amt für Bodendenkmalpflege im Rheinland Das Ganze ist mehr als die Summe seiner Teile Das Ganze ist mehr als die Summe seiner Teile. Festschrift für Jürgen Kunow, 2018
The aim of this contribution is on the one hand to map pre-industrial long distance roads located... more The aim of this contribution is on the one hand to map pre-industrial long distance roads located in a hilly region east of Cologne, Germany as exactly as possible and on the other hand to assess the accuracy of least-cost approaches that are increasingly applied by archaeologists for prehistoric road reconstruction. Probably the earliest map covering the study area east of Cologne dates back to 1575. The map is distorted so that rectification is difficult. But it is possible to assess the local accuracy of the map and to transfer the approximate routes to a modern map manually. Most of the area covered by the 1575 map is also depicted on a set of more accurate maps created in the early 19th century and a somewhat later historical map set (ca. 1842 AD). The historical roads on these rectified historical maps close to the approximate roads were digitized and compared to the outcomes of least-cost analysis, specifically least-cost paths and accessibility maps. Based on these route reconstructions with limited accuracy, Lidar data is checked to identify remains of these roads. Several approaches for visualizing Lidar data are tested to identify appropriate methods for detecting sunken roads. Possible sunken roads detected on the Lidar images were validated by checking cross sections in the digital elevation model and in the field.
For part of Bergisches Land east of Cologne, Germany, the maps by Ploennies finished in 1715 AD a... more For part of Bergisches Land east of Cologne, Germany, the maps by Ploennies finished in 1715 AD and the accompanying text provide detailed data for testing approaches used in archaeology for determining the boundaries of agrarian production. The historical maps indicate the settlement sizes ranging from single farmsteads to towns with walls. In 1715 AD the towns were quite small and the economy in this rural area relied mostly on agriculture (oats and livestock). The boundaries of agrarian production can be estimated by applying methods for allocating farmland territories to the settlement locations. The methods discussed are site catchments, retrodictive modelling for neighbourhoods, Thiessen polygons and approaches based on the distances between neighbouring settlement locations. With the exception of retrodictive modelling for neighbourhoods, two variants of the approaches are applicable, either resting on straight-line or least-cost distances.
Digital elevation models (DEMs) are an important basis for many archaeological GIS studies like p... more Digital elevation models (DEMs) are an important basis for many archaeological GIS studies like predictive modelling, visibility and least-cost path analysis. Issues in DEM data have been discussed in textbooks dealing with archaeological GIS applications and in some papers published more than ten years ago. But lately hardly any archaeological GIS study relying on DEM data refer to these issues though the problems connected with DEM use have not decreased with increasing availability of DEM data. This contribution presents a case study in the East Andean slopes of Ecuador analysing the impact of DEM choice on slope and aspect calculation as well as least-cost site catchments and least-cost paths. Four freely available elevation data sets are considered: ASTER GDEM2, SRTM at 3 arc-second and 1 arc-second resolution as well as a DEM derived from digitised contour lines of the official 1:50,000 maps provided by the Military Geographical Institute of Ecuador. Issues discussed are DEM resolution, horizontal and vertical accuracy, filling voids, and creating a DEM from contour lines.
CAA 2015 - Keep the Revolution Going. Proceedings of the 43rd Annual Conference on Computer Applications and Quantitative Methods in Archaeology held in Siena 2014
People who walk into unknown territory see only part of the landscape ahead whereas optimal path ... more People who walk into unknown territory see only part of the landscape ahead whereas optimal path calculation relies on the assumption that the walker has total knowledge of the landscape ahead. For this reason optimal path calculations are not appropriate for modelling dispersal processes. But the cost function derived from optimal path calculations can also be applied for modelling dispersal: The walker will prefer paths consisting of locally optimal steps in the direction chosen initially. Based on this assumption, a new agent-based algorithm is proposed for modelling the dispersal into unknown terrain starting from a given location. The corridors of movement generated this way are compared to globally optimal radial paths for a hilly study area located east of Cologne in Germany.
Recently, reconstructing ancient paths has become a standard tool in archaeological spatial analy... more Recently, reconstructing ancient paths has become a standard tool in archaeological spatial analysis for non-uniform terrain. In most applications, the route reconstruction is based on Dijkstra's algorithm that generates the globally optimal path between two locations. Fabrega Alvarez. and Parcero Oubina (2007) proposed
an algorithm that identifies optimal radial paths for a given location. These globally optimal paths require full knowledge of the surroundings, but a newcomer will proceed differently. A new agent-based model is presented that identifies probable
paths for exploring the unknown terrain by a newcomer who has only limited knowledge of the landscape ahead. This contribution will compare the outcomes of newcomer models with those of full knowledge models for a study area with several
known ancient routes in the Bergisches Land, Germany.
An early application of GIS methods in archaeology is the test of the hypothesis that sites of so... more An early application of GIS methods in archaeology is the test of the hypothesis that sites of some period and type are associated with roads of the same or another period. A case study in the Bergisches Land, Germany raised several issues: (i) In most archaeological applications, the one-sample Kolmogorov Smirnov test is based on the straight-line distance, but least-cost distances are often more appropriate. (ii) Is the site sample biased? (iii) What is the impact of the sites’ autocorrelation on the test? (iv) In a network of roads, how does each road contribute to the test result?
Der Gletscher Chimborazo ist mit ca. 6280 m ü.NN der höchste Berg Ecuadors. Unter anderem ist die... more Der Gletscher Chimborazo ist mit ca. 6280 m ü.NN der höchste Berg Ecuadors. Unter anderem ist dieser Berg dadurch bekannt, dass Alexander von Humboldt versuchte, den Gipfel zu erreichen, doch bei einer Höhe von ca. 5600 m ü.NN aufgeben musste. Im Jahr 2012 führte Alden Yépez an der Südflanke des Berges eine Prospektionsmaßnahme durch, um Fundstellen zu dokumentieren, die durch Raubgrabungen bedroht sind. Dies erbrachte Erkenntnisse zu zwölf Fundstellen auf einer Höhe zwischen 3400 und 4900 m ü.NN. Dabei handelt es sich vermutlich bei einem Teil der Fundstellen um Kultstätten der Inka, die in Zusammenhang mit Menschenopfern zu sehen sind. Diese Fundstellen orientieren sich weitgehend an einem Altweg. Streckenweise ist dieser Altweg auch heute noch begehbar, so dass er auch im Zuge der archäologischen Untersuchung mit einem GPS dokumentiert werden konnte.
Eine räumliche Analyse im Prospektionsgebiet geht der Frage nach, ob sich dieser Altweg durch entsprechende Rechenmethoden rekonstruieren lässt. Zum Vergleich werden auch spätere Wege in der Umgebung des Chimborazo herangezogen, z.B. Schmugglerpfade. Eine Erreichbarkeitskarte gibt Aufschluss darüber, ob die Fundstellen sich an Orten hoher Erreichbarkeit befinden. Dabei fällt auf, dass drei dicht beieinander liegende Fundplätze in schlecht erreichbarem Gebiet liegen, während andere, noch höher gelegene Fundplätze besser erreichbar sind. Deshalb wird die Frage untersucht, welche anderen Faktoren bei der Standortwahl eine Rolle spielten.
The trade route Brüderstraße, which had been used probably since early medieval times, connected ... more The trade route Brüderstraße, which had been used probably since early medieval times, connected Cologne with the mining sites in the Siegen area about 80 km east of Cologne. The small town Wildberg is located on this route. On modern maps, a silver mine is depicted close to Wildberg. The silver mines in this area were in exploited already in medieval times. According to historical sources, since about 1250 AD coins were manufactured in Wildberg, but in 1275 King Rudolf allowed Count Adolf of Berg to move the mint to Wipperfürth. So the silver had to be transported for at least 33 km, if the shortest path – as the crow flies – was taken. However, the known medieval trade routes in this region are not as ruler-straight as many sections of Roman roads. This paper addresses the question if it was efficient to use the known old trade routes for transporting the silver from Wildberg to Wipperfürth. In the 16th century the silver produced in Wildberg was mainly transported to Cologne, most probably on the Brüderstraße. But Wipperfürth is located north of the Brüderstraße, so the Brüderstraße is no option for connecting Wildberg with Wipperfürth. The study of the route layout of the Brüderstraße helps to reconstruct a probable medieval transport route from Wildberg to Wipperfürth. Several route layout models were tested by calculating least-cost paths. The least-cost path which is closest to the Brüderstraße is probably based on the best model. None of the models is perfect which is partly due to inaccuracies of the data employed. However, the best least-cost path is close enough to the Brüderstraße, so that that a plausible cost model for medieval transport in the study region is derived from the least-cost path study. This cost model is applied to construct the most efficient route from Wildberg to Wipperfürth. A substantial part of the least-cost path between Wildberg and Wipperfürth coincides with the old trade route Bergische Eisenstraße, which is said to date back to the late middle ages, i.e. the time when silver was transported from Wildberg to the mint in Wipperfürth.
Several models for least-cost networks are available:
(i) all pair shortest paths
Within an all... more Several models for least-cost networks are available:
(i) all pair shortest paths
Within an all pair shortest paths network, each pair of contemporary sites in a region is connected by a least-cost path. According to this model, whenever a new site was founded within the region considered, this site established shortest path connections to all other sites. The number of connections required has a quadratic growth rate, and this is one of the reasons why in general such networks consist of a limited number of sites only. Such networks are highly redundant and without centrality. The relative chronology of the paths cannot be deduced from the path layout.
(ii) Least-cost network to the builder
A least-cost network to the builder connects all sites but minimizes the total cost of route construction. All site locations must be known for creating this network. Consequently, a chronological development of this network is not part of the model. In a landscape with constant friction, Steiner trees solve this problem, however, calculation complexity is high. Network models that are a compromise between models (i) and (ii) have also been developed.
(iii) Triangulation network
Triangulation networks connect each site with its immediate neighbours by least-cost paths which form a Delaunay triangulation in a landscape with constant friction. If most interaction takes place between neighbours, this network is very efficient. Such networks are with hardly any centrality. In a Delaunay triangulation, adding a new point can change many triangles in the network; however, this is unlikely to happen when a new site is added to an existing network.
(iv) Central point network
Central point networks spread from a single point, thus exhibiting a high degree of centrality. Based on the network structure, it does not seem possible to reconstruct the chronological sequence of the routes in such a network.
(v) Main travel route with subsidiary paths
The central feature of a network could also be a line rather than a point, e.g. a river or a main travel route. For a central line, least-cost subsidiary paths can be constructed, forming yet another network configuration. It is quite obvious that the central line must have existed before the subsidiary paths were created, however, the chronological sequence of the subsidiary paths is not inherent in the path layout.
The paper will provide a synthesis of these models, some of which have already been discussed in archaeological least-cost studies. Unfortunately, most of these models are not supported in GIS software and many archaeologists using the software are not aware of the model implemented in the procedure they apply in their study.
The Bergisches Land in Germany serves as a test landscape for these models. This region provides less attractive natural conditions than the Rhine area and for this reason, only very few settlements existed in this region prior to the year 1000 AD. Detailed historical sources and old maps are available to check if the models agree with the historical reality.
Accessibility to fresh water, good soil, salt and raw material like flint stones played an import... more Accessibility to fresh water, good soil, salt and raw material like flint stones played an important role in prehistory. For example, accessibility to water is an important factor included in most archaeological predictive models. Several definitions for the geographical term accessibility exist. Accessibility is often determined in relation to target locations, for example accessibility to fresh water supply. The geometry of the target locations are often points (e.g. sources) but could also be lines (e.g. streams). The simplest approach to calculate accessibility is based on the presence or absence of a target, more refined methods take several targets and their weights into account, for example the amount of water available at a fresh water supply. In an archaeological context, accessibility often is considered as the ease with which a location may be reached from any other location in the area. The term general accessibility will be used in this context. Some studies assume that an area was explored by the first settlers using corridors of high accessibility. It will be shown that for archaeological landscape analysis, popular accessibility measures derived from the gravity model are computationally expensive and may produce unintuitive results. Instead this paper presents a more intuitive method for calculating accessibility which is inspired by kernel density estimation. The kernels take the topographic situation into account, i.e. least-cost distances are calculated instead of Euclidian distances. General accessibility depends on the scale of the analysis, a location which is highly accessible on a local scale may hardly be reachable on a more global scale. The bandwidth parameter of kernel density estimation can be used to control the scale of the accessibility measure. The Bergisches Land in Germany provides the test region where the concepts and methods presented in this paper will be applied.
Die Aufsiedlung des Bergischen Landes fand erst im 11.-13. Jh. statt, für eine Vielzahl von Orten... more Die Aufsiedlung des Bergischen Landes fand erst im 11.-13. Jh. statt, für eine Vielzahl von Orten ist die Erstnennung dokumentiert, und auch der Verlauf der mittelalterlichen Fernstraßen ist bekannt. Somit bietet dieser Raum die Möglichkeit, Methoden zur Rekonstruktion von Altwegen und zur Analyse von Siedlungsstrukturen zu testen. Auch im 137,5 km² großen Untersuchungsgebiet Quijos und Cosanga am Ostabhang der Anden sind Abschnitte von Altwegen bekannt. Ziel der räumlichen Analyse war ursprünglich, die Bewegungsmuster zu rekonstruieren und auf dieser Grundlage die Prospektionsdaten auszuwerten. Es zeigte sich jedoch, dass die Prospektionsdaten sehr problematisch sind. Dennoch erlaubten es die Wegerekonstruktionen, Aussagen über die Siedlungen der letzten vorspanischen Phase zu treffen.
Outline: The main objectives are: (i) to analyse the conditions that must be fulfilled for a succ... more Outline: The main objectives are: (i) to analyse the conditions that must be fulfilled for a successful density reconstruction of settlement patterns based on survey data and (ii) to present a new method for creating density maps that takes the friction of the landscape into account.
Abstract: At a regional scale, find density analysis based on survey data is often applied to reconstruct settlement patterning. However, this is by no means a trivial task because the finds recovered during the survey often are not a random sample of the initial distribution. Moreover the density estimation method plays an important role. A check list is presented with conditions that must be fulfilled for a successful density reconstruction of settlement patterns based on finds. Moreover, a new method for creating density maps is suggested which takes the friction of the landscape into account. Density methods based on distances derived from straight-line connections are only appropriate in landscapes with low variation in friction. However in regions with steep slopes, boggy terrain or streams that cannot be traversed easily, map distances often grossly underestimate the true effort required to transport goods from A to B. Least-cost path analysis has been applied successfully in some archaeological studies to reconstruct prehistoric and historic movement patterns. As human movement patterns of the past largely determine the distribution of artefacts, find density analysis should take these patterns into account. This is performed by modifying the kernel density method. The test case for the ideas presented above are shovel probe and surface collection data from an East Andean region in Ecuador.
Keywords: survey, GIS, settlement patterns, kernel density estimation, least-cost paths
Vor knapp zehn Jahren startete ein Prospektionsprojekt am Osthang der Anden, im Gebiet der Flüsse... more Vor knapp zehn Jahren startete ein Prospektionsprojekt am Osthang der Anden, im Gebiet der Flüsse Quijos und Cosanga in Ecuador. Ziel der Prospektion war es, für dieses 137,5 km² große Untersuchungsgebiet die Besiedlungsgeschichte zu rekonstruieren und damit auch die sozialen Prozesse, insbesondere die Entwicklung von Häuptlingstümern. Die häufig sehr dichte Vegetation in diesem topographisch stark bewegten Gebiet erlaubte eine mehr oder weniger systematische Aufsammlung von Funden nur auf einem geringen Teil der untersuchten Flächen. Andere Flächen wurden durch so genannte Schaufelproben untersucht, das sind würfelförmige Erdaushübe mit einer Kantenlänge von ca. 60 cm. Insgesamt konnten 24024 Scherben auf 2154 Flächen dokumentiert werden. Hinzu kommen Funde aus 46 Sondagen (1x1 bzw. 1x2 m), sie bildeten die Grundlage eines relativen Chronologieschemas, das vier Perioden unterscheidet, wobei ein Großteil der Funde der letzten noch nicht von spanischer Besetzung gekennzeichneten Periode zuzuordnen ist. Die so vorliegenden Daten wurden von den Autoren quellenkritisch untersucht. Für die Auswertung der Daten wird in der Regel eine Zufallsstichprobe voraus gesetzt. Erste textliche Hinweise auf systematische Verzerrungen fanden sich in der Publikation der Prospektionsdaten. Daraufhin haben wir angefangen, die Quellenfilter zu quantifizieren und nach Korrekturmöglichkeiten gesucht. So ist die geologische Situation im Untersuchungsgebiet durch Erdrutsche und Überschwemmungen kontinuierlichen Veränderungen ausgesetzt. Deshalb versuchten wir, aufgrund des Abrollungsgrades der Scherben und dem Anteil von Feinkeramik durch Verlagerungsprozesse geprägte Flächen zu identifizieren. Außerdem untersuchten wir, ob Funde aus Schaufelproben und Aufsammlungen vergleichbare Verteilungen aufweisen. Auf der Grundlage der aus diesen Arbeitsschritten gewonnenen Erkenntnisse werden Methoden vorgestellt, um die Prospektionsdaten so zu transformieren, dass sie stärker einer Zufallsstichprobe gleichen.
Abstract. Parallel coordinate plots (PCPs) can be applied to explore multivariate data with more ... more Abstract. Parallel coordinate plots (PCPs) can be applied to explore multivariate data with more than three dimensions. This visualisation method is straightforward and easily intelligible for people without statistical background. However, to our knowledge, PCPs have not yet been applied in archaeology. This paper will present some examples of archaeological classifications which are clearly visible in PCPs. For this purpose a program has been written which offers some additional options which are not supported in standard software for PCP generation. Some of the functionality of Geographic Information Systems (GIS) was introduced for PCPs: This program is able to create a thematic display based on a user-selected variable, optionally multiple plots highlight each thematic colour. Another variable may control the breadth of the PCP lines. Moreover, an info-tool, zoom, and a find-function are supported. The resulting graph can be saved in SVG format and in a GIS format.
Least-cost path (LCP) calculations consist of several steps, and require several choices: First t... more Least-cost path (LCP) calculations consist of several steps, and require several choices: First the cost components should be identified, the next step is the selection of an appropriate cost function for each component and a method for combining these functions. The result depends also on the number of neighbours considered in the raster grid, the approach implemented for anisotropic cost calculations, and the implementation of the LCP algorithm. Moreover, the accuracy of the environmental data plays a role. For routes showing evidence of planning and purposeful construction like roadbeds, pavement or bridges other approaches are required than for roads that more or less built and maintained themselves. It is recommended to start with a simple model and to refine this model successively. Due to the uncertainties regarding the cost components and the reliability of the geographic data, the stability of the results should be investigated, for example by varying the contribution of the different cost components to the cost function. Several models for least-cost networks exist: The most simple reconstruction of a least-cost network is by repeated application of a method for calculating a single LCP. This is only appropriate, if the aim is to minimise the travel costs between any two locations in the network. Another objective might be to optimise the road network so that it covers only a minimum amount of space of the valuable land used for agrarian purposes. Still another model might focus on the relative chronology of the routes which formed a network at some point in time or the origin of exploration for the study area considered. However, most archaeological network reconstruction studies do not explicitly discuss these alternatives but rely on the solution generated by an off-the-shelf least-cost network implementation.
Die 'Arbeitsgemeinschaft Computeranwendungen und Quantitative Methoden in der Archäologie' (AG CA... more Die 'Arbeitsgemeinschaft Computeranwendungen und Quantitative Methoden in der Archäologie' (AG CAA) veranstaltet jährlich einen gleichnamigen Workshop, der 2013 an der Freien Universität Berlin stattfand. Die Aufsätze zeigen, dass dreidimensionale Daten bei der archäologischen Dokumentation und Analyse immer wichtiger werden. 3D-Modelle oder digitale Geländemodelle (DGMs) lassen sich mithilfe von photogrammetrischen Methoden (Structure-from-Motion) oder Laserscanning erstellen. Über die Dokumentation von Funden und Befunden hinaus erlauben solche Daten zum Beispiel die virtuelle Rekonstruktion von eingestürzten Mauern, die Mustererkennung auf Keilschrifttafeln oder auch die Analyse der Beleuchtung im Innern eines griechischen Tempels. Archäologische Befunde werden in hoch aufgelösten DGMs (Lidar-Daten) sichtbar. DGMs sind auch die Basis für Sichtbarkeitsanalysen und Berechnungen zur Wegerekonstruktion. Diese Publikation ausgewählter Beiträge zu aktuellen Dokumentations- und Analyseverfahren in den Altertumswissenschaften deckt einen weiten Bereich typischer CAA-Themen ab.--- In der Reihe BERLIN STUDIES OF THE ANCIENT WORLD erscheinen Monographien und Sammelbände aller altertumswissenschaftlichen Disziplinen. Die Publikationen gehen aus der Arbeit des Exzellenzclusters "Topoi. The Formation and Transformation of Space and Knowledge in Ancient Civilizations" hervor, einem Forschungsverbund der Freien Universität Berlin und der Humboldt-Universität zu Berlin sowie den Partnerinstitutionen Berlin-Brandenburgische Akademie der Wissenschaften, Deutsches Archäologisches Institut, Max-Planck- Institut für Wissenschaftsgeschichte und Stiftung Preußischer Kulturbesitz. Die Reihe ist Bestandteil der Publikationsplattform Edition Topoi. Alle Bände der Reihe sind elektronisch unter www.edition- topoi.org verfügbar.
This session invites papers that discuss applications of GIS-based visibility analysis in cultura... more This session invites papers that discuss applications of GIS-based visibility analysis in cultural landscapes and built environments. We would especially like to encourage works on new approaches to the study of visual space and contributions that aim to bring to the front theoretical aspects of human visual perception in cultural heritage sites. Presentations exploring the visual properties of cultural spaces using other methods, such as space syntax (e.g. 2D or 3D isovists, visibility graphs) and formal lighting analysis, are also welcome.
In 1980 Orton discussed how to include information about finds in a Harris diagram. One suggestio... more In 1980 Orton discussed how to include information about finds in a Harris diagram. One suggestion was to include symbols for the presence or absence of finds in the diagram. This is quite easy if only four find types are considered as in Orton's example. Since the presence of a type is more significant than its absence, it appears to be sensible to include symbols solely for the presence of types. An example which was created by a new version of my program HARRIS (Herzog & Scollar 1991, Herzog 1993) is given in Fig. 16.1. Another presentation of finds is as a table resulting fi-om seriation. All the strata on a single level in the Harris diagram may be merged and presented as one feature in the table. Alternatively, the strata within a level could be ordered arbitrarily and all strata containing finds shown in the table. The latter presentation was added to the HARRIS program. But finds help us with dating a stratum, especially if a stratum belongs to a so-called floating seque...
What are the issues stratigraphic computing has to deal with? This is a personal view presenting ... more What are the issues stratigraphic computing has to deal with? This is a personal view presenting the features of my program Stratify and highlighting trends and aspects that have been discussed in recent publications. Each Harris diagram program has to store its data, either in a data base, or in dynamic structures like linked lists or resizeable arrays. Data storage is the basis of all stratigraphic computing, and with a data base, it is easy to perform simple queries, copy operations, etc.. Of course, the Harris diagram software must create a layout, it is desirable that it minimises crossings, and is capable of dealing with groups and phases. One may want to enter data in one’s own data base and import it into the Harris diagram program later, and probably export the layout results. In my opinion, the most important aspect in Harris diagram computing is dating: The chronological relationships do not result in an unambiguous relative chronology, and therefore additional informatio...
Zusammenfassung-Innerhalb eines EU-Projektes erarbeitet eine Arbeitsgruppe der Abteilung Prospekt... more Zusammenfassung-Innerhalb eines EU-Projektes erarbeitet eine Arbeitsgruppe der Abteilung Prospektion des Rheinischen Amtes für Bod end enkmalpflege (Lan d schaftsverban d Rheinlan d) für d en Flächennutzungsplan d er Stad t Erkelenz, Kreis Heinsberg, einen Fachbeitrag mit d en bis d ato bekannten archäologischen Fund plätzen. Ziel ist es, Archäologie und historische Kulturlan d schaftsrelikte frühzeitig im Planungsprozeß zu integrieren und hierfür eine Strategie zu entwickeln. Ein wesentliches Arbeitsinstrument in diesem Prozeß stellt d ie Zentrale Archäologische Datenbank (kurz ZADAB) d ar. Hier werd en in verschied enen Datenbankmod ulen d ie archäologischen Informationen d es Amtes verwaltet. Dazu gehören d iverse archäologische Aktivitäten wie Zufallsmeldungen, Prospektionen, Ausgrabungen, Luftbild er, aber auch Bodendenkmälerund archäologische Plätze. Alle projektrelevanten Daten, von d en Fund stellen über d ie Kulturland schaftsrelikte, d en Luftbild ern bis zu d en Störzonen...
The Quijos and Cosanga study area is located in the East Andean mountains in Ecuador. This area i... more The Quijos and Cosanga study area is located in the East Andean mountains in Ecuador. This area is probably the origin of the Cosanga pottery which has also been recorded in the late cultures of the Andes in northern Ecuador. The trade of pottery in prehispanic time is one of the reasons for studying this area in terms of movement patterns. Based on the assumption that the ancient movement patterns are preserved in old paths, least-cost path analysis is used to identify the main friction factors. Four old paths are considered, and it seems that slope and crossing streams are important friction factors, though the match of the known routes and the least-cost path reconstructions is not perfect in most cases. Based on the model describing the movement patterns, the routes between known ancient path segments are reconstructed. Moreover, least-cost site catchments are generated for seven Late Period settlements recorded in a survey project directed by Andrea Cuéllar in 2002. According t...
The aim of this contribution is on the one hand to map pre-industrial long distance roads located... more The aim of this contribution is on the one hand to map pre-industrial long distance roads located in a hilly region east of Cologne, Germany, as exactly as possible and on the other hand to assess the accuracy of least-cost approaches that are increasingly applied by archaeologists for prehistoric road reconstruction. Probably the earliest map covering the study area east of Cologne dates back to 1575. The map is distorted so that rectification is difficult. But it is possible to assess the local accuracy of the map and to transfer the approximate routes to a modern map manually. Most of the area covered by the 1575 map is also depicted on a set of more accurate maps created in the early 19th century and a somewhat later historical map set (ca. 1842 AD). The historical roads on these rectified historical maps close to the approximate roads were digitized and compared to the outcomes of least-cost analysis, specifically least-cost paths and accessibility maps. Based on these route re...
According to a historical document dating back to 1065 AD, the medieval road known as strata Colo... more According to a historical document dating back to 1065 AD, the medieval road known as strata Coloniensis connected the monastery of Essen-Werden with Cologne in Germany. Several alternative hypotheses concerning the course of this route were published, they only agree on a fairly small section south of Essen-Werden. The aim of the paper is to identify the most probable course of the medieval road between Essen-Werden and the river Düssel. In their arguments, the supporters of two hypotheses refer to historical maps. After outlining the hypotheses, the paper discusses these historical maps and additional earlier maps, several of them are available in online repositories. For each of the early maps considered, the roads or paths depicted were digitized, mostly on the basis of place names and water courses that could be found both on the old and a more recent map. The results allow assessing the accuracy of the maps and investigating the continuity of the roads. Visualization of high r...
Since 2012, field walking surveys and geophysical prospection as well as an excavation have docum... more Since 2012, field walking surveys and geophysical prospection as well as an excavation have documented remains of several Sue pottery kilns dating in the 9 century at Nakadake Sanroku, Kagoshima prefecture, southern Japan (Fig. 1a; Nakamura and Yoshimoto, 2015; Matsusaki, 2018). They were constructed by digging tunnels into the slope of hills, typically with a width of about 2 m and a length of more than 5 m. Most of the tunnels collapsed in the course of time forming elongated depressions with a depth of about 20 cm. These remains are hard to detect in the densely forested, mountainous area (Fig. 1c). Therefore, airborne Lidar was considered the method of choice for effectively identifying additional kiln sites. Only a few Lidar projects have been carried out in Japan up to now, and these focused on more prominent features such as castle remains. But this new project had not only to deal with the dense vegetation and large differences in altitude, but also with mostly inconspicuous...
National or regional archaeological databases have a long tradition in digital cultural heritage ... more National or regional archaeological databases have a long tradition in digital cultural heritage management. For instance, the book edited by Larsen in 1992 provides information on such databases in nine different countries, including the database in the Rhineland Commission for Archaeological Sites and Monuments (abbreviation in German: ABR) in Bonn, Germany (Scollar, 1992). As in most countries, the software for this database has been updated several times since 1992, and this information system nowadays includes a powerful GIS component as well as a document management system (Herzog et al. 2014; Herzog 2018). An important component of this system called BODEON is the management of planning projects, and this paper will focus on this unique component.
National or regional archaeological databases have a long tradition in digital cultural heritage ... more National or regional archaeological databases have a long tradition in digital cultural heritage management. For instance, the book edited by Larsen in 1992 provides information on such databases in nine different countries, including the database in the Rhineland Commission for Archaeological Sites and Monuments (abbreviation in German: ABR) in Bonn, Germany (Scollar, 1992). As in most countries, the software for this database has been updated several times since 1992, and this information system nowadays includes a powerful GIS component as well as a document management system (Herzog et al. 2014; Herzog 2018). An important component of this system called BODEON is the management of planning projects, and this paper will focus on this unique component.
An important and frequently used tool in archaeological spatial analysis is least-cost path (LCP)... more An important and frequently used tool in archaeological spatial analysis is least-cost path (LCP) analysis to compute routes connecting a set of targets. The outcome depends on the cost model chosen and the topographic data used. A slope-dependent cost model requires a digital elevation model (DEM) that should reflect the landscape in the past. It is often impossible to reconstruct the past terrain, and modern high-resolution elevation data results in problematic storage requirements and computation times. This article presents a case study that explores issues in replication and stability of LCP calculations for pairs of targets that are close to known old trade routes. A large number of cost models is tested based on two topographic data sets, including DEMs of two different resolutions (25 m and 50 m). The cost models use six different slope-dependent cost functions suggested by various authors for pedestrian movement. Moreover, a slope-dependent cost function is applied that res...
Uploads
Papers by Irmela Herzog
Archaeological sites are increasingly threatened by climate-related hazards. In response, heritage management authorities initiated projects to document damage and plan risk assessment measures. We present a project initiated after the heavy rainfall and subsequent flood event of July 2021, which involved extensive fieldwork to document the damage to archaeological sites in the Rhineland. We use this database to characterise and assess the damage and investigate site-specific and geospatial factors to identify potential predictive parameters for site damage.
Results
During fieldwork, we found that the flood damaged 19% of the 538 archaeological sites surveyed. The majority of damaged sites are relatively recent, dating from the medieval or modern periods, and are associated with the use of water power. Damage was mainly caused by erosion, floating debris and washout, e.g. mortar. In a case study, we tested the option of comparing pre- and post-disaster Airborne Laser Scanning elevation data to identify damages. It showed that not only the damage detected during fieldwork was found but also additional areas of loss. In general, however, and quantified based on the entire dataset, the ordnance survey Airborne Laser Scanning data were of limited use for monitoring flood-related damage and could not replace fieldwork. Our statistical analysis of possible risk factors, including both site characteristics and geospatial parameters, using Naïve Bayes Modelling and chi-squared tests, showed that no set of parameters could consistently predict the preservation or damage of archaeological sites across all catchments. In contrast, some external geospatial factors correlated with the occurrence of damage.
Conclusions
The study highlights both the strengths and limitations of the approaches used to assess and predict the damage to the archaeological heritage in the 2021 flood zones of the Rhineland. It also demonstrates the complexity of the data and spatial processes involved, which limits generalisation but can still inform decision-making for archaeological site management and on-site protection measures in flood-prone areas. With the prospect of more frequent heavy rainfall due to climate change, the specific needs of the archaeological heritage should be integrated into broader prevention and disaster management plans.
LiDAR-Daten bestehen aus unregelmäßig verteilten Höhenpunkten auf der Erdoberfläche. Damit erlauben sie im Gegensatz zu Luftbildern auch in Waldgebieten einen von Bäumen ungestörten Blick auf die dort befindlichen Bodendenkmäler. Diese sind durch die Bewaldung teilweise noch obertägig erhalten, aber keineswegs geschützt: Rückegassen der Waldwirtschaft, Mountainbike-Routen, aber auch tief wurzelnde Bäume gefährden ihren Erhalt. Deshalb ist die Auswertung von LiDAR-Daten in Waldgebieten für das LVR-Amt für Bodendenkmalpflege im Rheinland besonders wichtig.
Archaeological sites are increasingly threatened by climate-related hazards. In response, heritage management authorities initiated projects to document damage and plan risk assessment measures. We present a project initiated after the heavy rainfall and subsequent flood event of July 2021, which involved extensive fieldwork to document the damage to archaeological sites in the Rhineland. We use this database to characterise and assess the damage and investigate site-specific and geospatial factors to identify potential predictive parameters for site damage.
Results
During fieldwork, we found that the flood damaged 19% of the 538 archaeological sites surveyed. The majority of damaged sites are relatively recent, dating from the medieval or modern periods, and are associated with the use of water power. Damage was mainly caused by erosion, floating debris and washout, e.g. mortar. In a case study, we tested the option of comparing pre- and post-disaster Airborne Laser Scanning elevation data to identify damages. It showed that not only the damage detected during fieldwork was found but also additional areas of loss. In general, however, and quantified based on the entire dataset, the ordnance survey Airborne Laser Scanning data were of limited use for monitoring flood-related damage and could not replace fieldwork. Our statistical analysis of possible risk factors, including both site characteristics and geospatial parameters, using Naïve Bayes Modelling and chi-squared tests, showed that no set of parameters could consistently predict the preservation or damage of archaeological sites across all catchments. In contrast, some external geospatial factors correlated with the occurrence of damage.
Conclusions
The study highlights both the strengths and limitations of the approaches used to assess and predict the damage to the archaeological heritage in the 2021 flood zones of the Rhineland. It also demonstrates the complexity of the data and spatial processes involved, which limits generalisation but can still inform decision-making for archaeological site management and on-site protection measures in flood-prone areas. With the prospect of more frequent heavy rainfall due to climate change, the specific needs of the archaeological heritage should be integrated into broader prevention and disaster management plans.
LiDAR-Daten bestehen aus unregelmäßig verteilten Höhenpunkten auf der Erdoberfläche. Damit erlauben sie im Gegensatz zu Luftbildern auch in Waldgebieten einen von Bäumen ungestörten Blick auf die dort befindlichen Bodendenkmäler. Diese sind durch die Bewaldung teilweise noch obertägig erhalten, aber keineswegs geschützt: Rückegassen der Waldwirtschaft, Mountainbike-Routen, aber auch tief wurzelnde Bäume gefährden ihren Erhalt. Deshalb ist die Auswertung von LiDAR-Daten in Waldgebieten für das LVR-Amt für Bodendenkmalpflege im Rheinland besonders wichtig.
an algorithm that identifies optimal radial paths for a given location. These globally optimal paths require full knowledge of the surroundings, but a newcomer will proceed differently. A new agent-based model is presented that identifies probable
paths for exploring the unknown terrain by a newcomer who has only limited knowledge of the landscape ahead. This contribution will compare the outcomes of newcomer models with those of full knowledge models for a study area with several
known ancient routes in the Bergisches Land, Germany.
Eine räumliche Analyse im Prospektionsgebiet geht der Frage nach, ob sich dieser Altweg durch entsprechende Rechenmethoden rekonstruieren lässt. Zum Vergleich werden auch spätere Wege in der Umgebung des Chimborazo herangezogen, z.B. Schmugglerpfade. Eine Erreichbarkeitskarte gibt Aufschluss darüber, ob die Fundstellen sich an Orten hoher Erreichbarkeit befinden. Dabei fällt auf, dass drei dicht beieinander liegende Fundplätze in schlecht erreichbarem Gebiet liegen, während andere, noch höher gelegene Fundplätze besser erreichbar sind. Deshalb wird die Frage untersucht, welche anderen Faktoren bei der Standortwahl eine Rolle spielten.
(i) all pair shortest paths
Within an all pair shortest paths network, each pair of contemporary sites in a region is connected by a least-cost path. According to this model, whenever a new site was founded within the region considered, this site established shortest path connections to all other sites. The number of connections required has a quadratic growth rate, and this is one of the reasons why in general such networks consist of a limited number of sites only. Such networks are highly redundant and without centrality. The relative chronology of the paths cannot be deduced from the path layout.
(ii) Least-cost network to the builder
A least-cost network to the builder connects all sites but minimizes the total cost of route construction. All site locations must be known for creating this network. Consequently, a chronological development of this network is not part of the model. In a landscape with constant friction, Steiner trees solve this problem, however, calculation complexity is high. Network models that are a compromise between models (i) and (ii) have also been developed.
(iii) Triangulation network
Triangulation networks connect each site with its immediate neighbours by least-cost paths which form a Delaunay triangulation in a landscape with constant friction. If most interaction takes place between neighbours, this network is very efficient. Such networks are with hardly any centrality. In a Delaunay triangulation, adding a new point can change many triangles in the network; however, this is unlikely to happen when a new site is added to an existing network.
(iv) Central point network
Central point networks spread from a single point, thus exhibiting a high degree of centrality. Based on the network structure, it does not seem possible to reconstruct the chronological sequence of the routes in such a network.
(v) Main travel route with subsidiary paths
The central feature of a network could also be a line rather than a point, e.g. a river or a main travel route. For a central line, least-cost subsidiary paths can be constructed, forming yet another network configuration. It is quite obvious that the central line must have existed before the subsidiary paths were created, however, the chronological sequence of the subsidiary paths is not inherent in the path layout.
The paper will provide a synthesis of these models, some of which have already been discussed in archaeological least-cost studies. Unfortunately, most of these models are not supported in GIS software and many archaeologists using the software are not aware of the model implemented in the procedure they apply in their study.
The Bergisches Land in Germany serves as a test landscape for these models. This region provides less attractive natural conditions than the Rhine area and for this reason, only very few settlements existed in this region prior to the year 1000 AD. Detailed historical sources and old maps are available to check if the models agree with the historical reality.
Several definitions for the geographical term accessibility exist. Accessibility is often determined in relation to target locations, for example accessibility to fresh water supply. The geometry of the target locations are often points (e.g. sources) but could also be lines (e.g. streams). The simplest approach to calculate accessibility is based on the presence or absence of a target, more refined methods take several targets and their weights into account, for example the amount of water available at a fresh water supply.
In an archaeological context, accessibility often is considered as the ease with which a location may be reached from any other location in the area. The term general accessibility will be used in this context. Some studies assume that an area was explored by the first settlers using corridors of high accessibility.
It will be shown that for archaeological landscape analysis, popular accessibility measures derived from the gravity model are computationally expensive and may produce unintuitive results. Instead this paper presents a more intuitive method for calculating accessibility which is inspired by kernel density estimation. The kernels take the topographic situation into account, i.e. least-cost distances are calculated instead of Euclidian distances.
General accessibility depends on the scale of the analysis, a location which is highly accessible on a local scale may hardly be reachable on a more global scale. The bandwidth parameter of kernel density estimation can be used to control the scale of the accessibility measure.
The Bergisches Land in Germany provides the test region where the concepts and methods presented in this paper will be applied.
Abstract: At a regional scale, find density analysis based on survey data is often applied to reconstruct settlement patterning. However, this is by no means a trivial task because the finds recovered during the survey often are not a random sample of the initial distribution. Moreover the density estimation method plays an important role. A check list is presented with conditions that must be fulfilled for a successful density reconstruction of settlement patterns based on finds. Moreover, a new method for creating density maps is suggested which takes the friction of the landscape into account. Density methods based on distances derived from straight-line connections are only appropriate in landscapes with low variation in friction. However in regions with steep slopes, boggy terrain or streams that cannot be traversed easily, map distances often grossly underestimate the true effort required to transport goods from A to B. Least-cost path analysis has been applied successfully in some archaeological studies to reconstruct prehistoric and historic movement patterns. As human movement patterns of the past largely determine the distribution of artefacts, find density analysis should take these patterns into account. This is performed by modifying the kernel density method. The test case for the ideas presented above are shovel probe and surface collection data from an East Andean region in Ecuador.
Keywords: survey, GIS, settlement patterns, kernel density estimation, least-cost paths
Die häufig sehr dichte Vegetation in diesem topographisch stark bewegten Gebiet erlaubte eine mehr oder weniger systematische Aufsammlung von Funden nur auf einem geringen Teil der untersuchten Flächen. Andere Flächen wurden durch so genannte Schaufelproben untersucht, das sind würfelförmige Erdaushübe mit einer Kantenlänge von ca. 60 cm. Insgesamt konnten 24024 Scherben auf 2154 Flächen dokumentiert werden. Hinzu kommen Funde aus 46 Sondagen (1x1 bzw. 1x2 m), sie bildeten die Grundlage eines relativen Chronologieschemas, das vier Perioden unterscheidet, wobei ein Großteil der Funde der letzten noch nicht von spanischer Besetzung gekennzeichneten Periode zuzuordnen ist.
Die so vorliegenden Daten wurden von den Autoren quellenkritisch untersucht. Für die Auswertung der Daten wird in der Regel eine Zufallsstichprobe voraus gesetzt. Erste textliche Hinweise auf systematische Verzerrungen fanden sich in der Publikation der Prospektionsdaten. Daraufhin haben wir angefangen, die Quellenfilter zu quantifizieren und nach Korrekturmöglichkeiten gesucht. So ist die geologische Situation im Untersuchungsgebiet durch Erdrutsche und Überschwemmungen kontinuierlichen Veränderungen ausgesetzt. Deshalb versuchten wir, aufgrund des Abrollungsgrades der Scherben und dem Anteil von Feinkeramik durch Verlagerungsprozesse geprägte Flächen zu identifizieren. Außerdem untersuchten wir, ob Funde aus Schaufelproben und Aufsammlungen vergleichbare Verteilungen aufweisen. Auf der Grundlage der aus diesen Arbeitsschritten gewonnenen Erkenntnisse werden Methoden vorgestellt, um die Prospektionsdaten so zu transformieren, dass sie stärker einer Zufallsstichprobe gleichen.
For this purpose a program has been written which offers some additional options which are not supported in standard software for PCP generation. Some of the functionality of Geographic Information Systems (GIS) was introduced for PCPs: This program is able to create a thematic display based on a user-selected variable, optionally multiple plots highlight each thematic colour. Another variable may control the breadth of the PCP lines. Moreover, an info-tool, zoom, and a find-function are supported. The resulting graph can be saved in SVG format and in a GIS format.
Several models for least-cost networks exist: The most simple reconstruction of a least-cost network is by repeated application of a method for calculating a single LCP. This is only appropriate, if the aim is to minimise the travel costs between any two locations in the network. Another objective might be to optimise the road network so that it covers only a minimum amount of space of the valuable land used for agrarian purposes. Still another model might focus on the relative chronology of the routes which formed a network at some point in time or the origin of exploration for the study area considered. However, most archaeological network reconstruction studies do not explicitly discuss these alternatives but rely on the solution generated by an off-the-shelf least-cost network implementation.
We would especially like to encourage works on new approaches to the study of visual space and contributions that aim to bring to the front theoretical aspects of human visual perception in cultural heritage sites. Presentations exploring the visual properties of cultural spaces using other methods, such as space syntax (e.g. 2D or 3D isovists, visibility graphs) and formal lighting analysis, are also welcome.