**2. Materials and methods**

The phenomena of gullying are affecting, on a global scale, large areas in the tropical and temperate regions [68-76]. In Romania, these are specific to the Moldavian Plateau, the Transylvanian Depression and the Getic Plateau [60-61, 65-66, 16, 77]. Apart from the archaeology and architecture applications, which are more visible to the public, the 3D ground laser scanning is successfully used in monitoring the processes of ravine-creation in eastern Romania. This technology was also used with promising results in monitoring the riverine valleys and the landslides [78, 5, 79, 38, 80-83]. As a result of its performance, it can be used successfully to assess the state of the environment, especially in land mapping and with limited extension in measuring the rate of erosion in certain land surfaces [33, 84-85].

**Figure 4.** The detailed cartographic representation of the upper sector or the gully

**Figure 3.** The final cartographic product for the investigated gully

and a greatly-diverse usability (Figure 4, 5).

several future papers.

surfaces [33, 84-85].

**2. Materials and methods** 

For a better monitoring of relief dynamics, the use of the GPS is imperiously necessary. This tool can delimit with accuracy the spatial limits of the phenomenon. These coordinates constitute the backbone of the morphological analysis of a dynamic land form. Unfortunately, the 3D scanner cannot precisely identify by itself the border between one area and another. By using the 3D scanner we are able to move to the next step of the analysis of the dynamics of the relief and man-made forms. The accuracy of the obtained data and the exceptional cartographical representation means that the 3D has a long use-life

For the study of the geomorphological processes we selected an area in the Moldavain Plateau that is affected by intense gullying. The complex analysis of the Cucuteni Ravine aims also at the impact it has on the Chalcolithic site. For all the other examples we presented only the fields of research they pertain to, while the results will be included in

The phenomena of gullying are affecting, on a global scale, large areas in the tropical and temperate regions [68-76]. In Romania, these are specific to the Moldavian Plateau, the Transylvanian Depression and the Getic Plateau [60-61, 65-66, 16, 77]. Apart from the archaeology and architecture applications, which are more visible to the public, the 3D ground laser scanning is successfully used in monitoring the processes of ravine-creation in eastern Romania. This technology was also used with promising results in monitoring the riverine valleys and the landslides [78, 5, 79, 38, 80-83]. As a result of its performance, it can be used successfully to assess the state of the environment, especially in land mapping and with limited extension in measuring the rate of erosion in certain land

**Figure 5.** The final cartographic representation of the gully (100 m grid)

In terms of definition, the 3D laser scanning is one of the new technologies, by which the geometry of an object or surface can be automatically measured, without the aid of a reflector, with speed and precision well above the ones of the classic solutions. The ground-

based laser scanner records the tridimensional points by measuring the vertical and horizontal angles, as well as the distance to each point. Even though there are largely different technologies, the field-use of the 3D scanner uses elements of the work methodology of the total station. In case the morphology of the scanned object is complex, more than one station points will be used. Thus, all its surfaces will be scanned. In case the shadowed surfaces are not reached by the laser beam, the software captures such areas automatically then integrates them to the scanned object. The process relies on registering distances and angles, and the data thus produced is used to compute the points' coordinates. The ability to register a massive amount of 3-D information in a relatively short time is the main advantage of this instrument, in contrast with classical equipment such as the total station [24, 10].

Use of Terrestrial 3D Laser Scanner in Cartographing and

Monitoring Relief Dynamics and Habitation Space from Various Historical Periods 55

external photographic camera can be attached to the scanner, we thought it to be no better

A particularly important phase of the fieldwork facet of the project was the geo-referencing of the point clouds. Using a reference station positioned on fixed known spot and a Leica 1200 GPS receiver, we referenced the point cloud established as the basis for the 3D model to the national coordinates system (Stereographic 1970). In fact, our methodology was based, for all of the three sessions, on computing the differences between the obtained georeferenced models using CAD and GIS. In respect to the raw data processing, it was carried out by filtering the data using the Cyclone dedicated software program, registering the data (see above), reducing the point cloud, creating a mesh by triangulation, and texturing the model. The final results of the analysis were produced by exporting sections, transverse and

As a case study, the selection of the Cucuteni-Baiceni Ravine for the present paper was based on its high level of activity and its lack of arboreal vegetation, which could have partially impeded the measuring of the volumetric parameters. The undergrowth vegetation had then to be erased using the technical methods allowed by the software. Furthermore, for increased accuracy, the edge of the gully was outlined during each session by using the two "traditional" instruments mentioned above (the reference station and the 1200 GPS receiver, both produced by Leica). The operation was somewhat cumbersome, because in such cases the data must be collected from extremely numerous positions, as to take into account all of the inflexions [44, 10]. All of the positions were geo-referenced and corroborated with older measurements. In this way, we were able to estimate the rate of soil erosion in the gully, for each of the measurements taken. The conjoint use of these two types of measurements (GPS

Another stage worth mentioning is, in the lab phase of the project, the filtering and modeling of data, to the aim of producing results compatible with the complementary software we used (CAD, GIS). Data filtering is a compulsory stage of the analysis, as the vegetation present at the scanning site, together with the very large amount of data, could result in errors in the Digital Terrain Model export-for-GIS process, as well as during the

longitudinal, of the three tridimensional models obtained in each session.

and 3D scanner) means that the risk of error was much diminished.

suited for the task than the internal camera [86].

**Figure 7.** The visual field of the scanner

The methodology used for the analysis of Cucuteni Ravine is strongly influenced by its impressive size and the local topography, which required, in the end, the use of 17 station points. To merge the 17 positions, 6 reflective tie-points (Figure 6) were used for each of them, except for the last one. Nevertheless, for the entire model we used 24 tie-points, since some of the scanning positions were referenced to tie-points also used by other station points, where the physical distance allowed. Terrestrial laser scanning (TLS) generates several point clouds, with local coordinates and additional info (the light intensity in the reflected beam, and the RGB values obtained from an external or internal photographic camera). The point clouds, after having been registered from different positions, must be merged as to obtain a complete model of the scanned target. This procedure is called "registration" and involves merging the point clouds through the use of reflective tie-points, specially built and delivered by the manufacturer, which are automatically recognized by the scanner when a very fine scan is performed.

**Figure 6.** The round target used as tie-point

For the current project, in all of the three scanning sessions, we employed a Leica ScanStation HDS 3600 3D scanner. It is a time-of-flight active scanner, which works by timing the round-trip time of a pulse of light. The operation range is 270º horizontally and 360º vertically (Figure 7), and the active distance is 300 m. With a resolution of 6 mm at a distance of 50 m, and due to its ability to register approximately 2000 points per second, the ScanStation HDS scanner is among the most effective equipment of its type. The average resolution for all of the scans was of approximately 6 mm, and the registered points numbered millions, despite the fact that the majority of positions overlapped. Although an external photographic camera can be attached to the scanner, we thought it to be no better suited for the task than the internal camera [86].

#### **Figure 7.** The visual field of the scanner

54 Cartography – A Tool for Spatial Analysis

station [24, 10].

the scanner when a very fine scan is performed.

**Figure 6.** The round target used as tie-point

based laser scanner records the tridimensional points by measuring the vertical and horizontal angles, as well as the distance to each point. Even though there are largely different technologies, the field-use of the 3D scanner uses elements of the work methodology of the total station. In case the morphology of the scanned object is complex, more than one station points will be used. Thus, all its surfaces will be scanned. In case the shadowed surfaces are not reached by the laser beam, the software captures such areas automatically then integrates them to the scanned object. The process relies on registering distances and angles, and the data thus produced is used to compute the points' coordinates. The ability to register a massive amount of 3-D information in a relatively short time is the main advantage of this instrument, in contrast with classical equipment such as the total

The methodology used for the analysis of Cucuteni Ravine is strongly influenced by its impressive size and the local topography, which required, in the end, the use of 17 station points. To merge the 17 positions, 6 reflective tie-points (Figure 6) were used for each of them, except for the last one. Nevertheless, for the entire model we used 24 tie-points, since some of the scanning positions were referenced to tie-points also used by other station points, where the physical distance allowed. Terrestrial laser scanning (TLS) generates several point clouds, with local coordinates and additional info (the light intensity in the reflected beam, and the RGB values obtained from an external or internal photographic camera). The point clouds, after having been registered from different positions, must be merged as to obtain a complete model of the scanned target. This procedure is called "registration" and involves merging the point clouds through the use of reflective tie-points, specially built and delivered by the manufacturer, which are automatically recognized by

For the current project, in all of the three scanning sessions, we employed a Leica ScanStation HDS 3600 3D scanner. It is a time-of-flight active scanner, which works by timing the round-trip time of a pulse of light. The operation range is 270º horizontally and 360º vertically (Figure 7), and the active distance is 300 m. With a resolution of 6 mm at a distance of 50 m, and due to its ability to register approximately 2000 points per second, the ScanStation HDS scanner is among the most effective equipment of its type. The average resolution for all of the scans was of approximately 6 mm, and the registered points numbered millions, despite the fact that the majority of positions overlapped. Although an A particularly important phase of the fieldwork facet of the project was the geo-referencing of the point clouds. Using a reference station positioned on fixed known spot and a Leica 1200 GPS receiver, we referenced the point cloud established as the basis for the 3D model to the national coordinates system (Stereographic 1970). In fact, our methodology was based, for all of the three sessions, on computing the differences between the obtained georeferenced models using CAD and GIS. In respect to the raw data processing, it was carried out by filtering the data using the Cyclone dedicated software program, registering the data (see above), reducing the point cloud, creating a mesh by triangulation, and texturing the model. The final results of the analysis were produced by exporting sections, transverse and longitudinal, of the three tridimensional models obtained in each session.

As a case study, the selection of the Cucuteni-Baiceni Ravine for the present paper was based on its high level of activity and its lack of arboreal vegetation, which could have partially impeded the measuring of the volumetric parameters. The undergrowth vegetation had then to be erased using the technical methods allowed by the software. Furthermore, for increased accuracy, the edge of the gully was outlined during each session by using the two "traditional" instruments mentioned above (the reference station and the 1200 GPS receiver, both produced by Leica). The operation was somewhat cumbersome, because in such cases the data must be collected from extremely numerous positions, as to take into account all of the inflexions [44, 10]. All of the positions were geo-referenced and corroborated with older measurements. In this way, we were able to estimate the rate of soil erosion in the gully, for each of the measurements taken. The conjoint use of these two types of measurements (GPS and 3D scanner) means that the risk of error was much diminished.

Another stage worth mentioning is, in the lab phase of the project, the filtering and modeling of data, to the aim of producing results compatible with the complementary software we used (CAD, GIS). Data filtering is a compulsory stage of the analysis, as the vegetation present at the scanning site, together with the very large amount of data, could result in errors in the Digital Terrain Model export-for-GIS process, as well as during the

volume calculation of the whole resulting model. In the first step, the tridimensional model was trimmed by removing the points lying outside the analyzed area, in order to reduce its originally large size. Concurrently, the model was scaled down in order to be more easily handled by the software. All these steps were carried out within the Cyclon platform, the proprietary software of the scanner, which was specially designed to handle the pointclouds in a 3D environment. The large number of axial and cross-sections of the model were produced at the same time, together with the various graphic, image and video exports.

Use of Terrestrial 3D Laser Scanner in Cartographing and

Monitoring Relief Dynamics and Habitation Space from Various Historical Periods 57

**Figure 9.** The 3D model and the map of the Cucuteni Ravine in 2008 (100 m grid)

**Figure 10.** The 3D model and the map of the Cucuteni Ravine in 2008 (10 m grid)

degree of generalization.

winter precipitation and high rainfall in spring. For the historical evolution of gullies, topographic maps and military plans of the Romanian Army were consulted. During the Second World War the Army had placed a battery of guns in the area of the unit studied. Unfortunately, it has only been possible to make use of zoning maps from after 1950. The maps drawn earlier are not accurate and they are often for orientation only, with a high

Meteorological data on precipitation, daily and monthly, were provided by the Meteorological Centre, Iasi, Moldavia. They were focused on Cotnari Meteorological Station, located near the Cucuteni-Baiceni ravines. The most important stations were rather uniformly distributed on Moldavian territory (Eastern Romania) [87]. Certain old

The stage of interpolating the scanner-produced points, to the aim of achieving a Digital Elevation Model, was carried out within the GIS platform (using the ArcGIS suite). Based on the resulting DEM, several derivates were produced (contour lines, volume calculations, various graphs and diagrams etc.) The proprietary software of Leica is also providing export functionality, but the options are more oriented toward GIS and CAD formats. In order to reduce to minimum the error margin of the volume calculations for the three models (of three successive years), which is the definitive element for outlining the timeline of change between the three scans, we used the Cyclon software. Following this procedure, there were no major, visible differences between the employed calculation methods.

The Cucuteni gully was selected for the present research because it is extremely active. The area occupied by the gully is very sparsely covered by vegetation, and the trees are virtually absent; therefore, nothing prevented volumetric measuring. The very sparse shrub vegetation was removed using the techniques made available by the dedicated software. The reason behind the selection of the gully for our investigation was due to the fact that the gullying process is affecting a very important archaeological site dating back to circa 5000 B.P. This made it easier to assess the rate of erosion over a period of great lengths.

Three consecutive measurements were performed at relatively equal intervals in 2008, 2009 and 2010 (Figure 8, 9, 10, 11). The last measurement was made in spring 2010 after a solid

**Figure 8.** The 3D scanner in action on the Cucuteni-Baiceni Ravine

Use of Terrestrial 3D Laser Scanner in Cartographing and Monitoring Relief Dynamics and Habitation Space from Various Historical Periods 57

**Figure 9.** The 3D model and the map of the Cucuteni Ravine in 2008 (100 m grid)

56 Cartography – A Tool for Spatial Analysis

volume calculation of the whole resulting model. In the first step, the tridimensional model was trimmed by removing the points lying outside the analyzed area, in order to reduce its originally large size. Concurrently, the model was scaled down in order to be more easily handled by the software. All these steps were carried out within the Cyclon platform, the proprietary software of the scanner, which was specially designed to handle the pointclouds in a 3D environment. The large number of axial and cross-sections of the model were produced at the same time, together with the various graphic, image and video exports.

The stage of interpolating the scanner-produced points, to the aim of achieving a Digital Elevation Model, was carried out within the GIS platform (using the ArcGIS suite). Based on the resulting DEM, several derivates were produced (contour lines, volume calculations, various graphs and diagrams etc.) The proprietary software of Leica is also providing export functionality, but the options are more oriented toward GIS and CAD formats. In order to reduce to minimum the error margin of the volume calculations for the three models (of three successive years), which is the definitive element for outlining the timeline of change between the three scans, we used the Cyclon software. Following this procedure, there were

The Cucuteni gully was selected for the present research because it is extremely active. The area occupied by the gully is very sparsely covered by vegetation, and the trees are virtually absent; therefore, nothing prevented volumetric measuring. The very sparse shrub vegetation was removed using the techniques made available by the dedicated software. The reason behind the selection of the gully for our investigation was due to the fact that the gullying process is affecting a very important archaeological site dating back to circa 5000

Three consecutive measurements were performed at relatively equal intervals in 2008, 2009 and 2010 (Figure 8, 9, 10, 11). The last measurement was made in spring 2010 after a solid

no major, visible differences between the employed calculation methods.

B.P. This made it easier to assess the rate of erosion over a period of great lengths.

**Figure 8.** The 3D scanner in action on the Cucuteni-Baiceni Ravine

**Figure 10.** The 3D model and the map of the Cucuteni Ravine in 2008 (10 m grid)

winter precipitation and high rainfall in spring. For the historical evolution of gullies, topographic maps and military plans of the Romanian Army were consulted. During the Second World War the Army had placed a battery of guns in the area of the unit studied. Unfortunately, it has only been possible to make use of zoning maps from after 1950. The maps drawn earlier are not accurate and they are often for orientation only, with a high degree of generalization.

Meteorological data on precipitation, daily and monthly, were provided by the Meteorological Centre, Iasi, Moldavia. They were focused on Cotnari Meteorological Station, located near the Cucuteni-Baiceni ravines. The most important stations were rather uniformly distributed on Moldavian territory (Eastern Romania) [87]. Certain old cartographic material was provided by the Military Topography Directorate in Bucharest. The orthophoto maps and the military maps dating from WW II were supplied by the ANCPI (The National Agency for Surveying and Real Estate Publicity) in.

Use of Terrestrial 3D Laser Scanner in Cartographing and

Monitoring Relief Dynamics and Habitation Space from Various Historical Periods 59

**Figure 12.** Making the morphologic map of the Cucuteni-Baiceni Ravine

measuring sessions (erosion only)

**Figure 13.** Cross-section through the Cucuteni-Baiceni Ravine, with the indication of the 2008 and 2010

**Figure 11.** The model used for volume calculation for the eroded and transported material (50 m grid)
