GISdevelopment.net --> Application --> Archaeology



Topographic information in cultural and natural heritage visualization and animation

Boehler W. , Scherer Y. , Siebold M. , Mainz
Institute for Spatial Information and Surveying Technology,
FH Mainz, University of Applied Sciences, Holzstrasse 36, 55116 Mainz, Germany,
Phone +49-6131-2628-27, Fax +49-6128-2628-15
E-Mail: i3mainz@geoinform.fh-mainz.de

Heinz, G.
Roemisch-Germanisches Zentralmuseum,
Forschungsinstitut für Vor- und Fruehgeschichte,
Ernst-Ludwig-Platz 2, 55116 Mainz, Germany


Abstract
Cultural heritage objects can only be understood, if the surrounding landscape is taken under consideration, too. In certain cases, the landscape even constitutes the cultural object itself. UNESCO has considered this fact by adding the category of 'Cultural Landscapes' to the sites eligible for the World Heritage List. In this paper, methods to survey, document and visualize landscapes are discussed. Especially, with new imaging opportunities, such as satellite images with 1 meter resolution, landscapes can be mapped economically. Results, such as maps, perspectives, animations and geographic information systems can help to present the landscapes to those who cannot visit its location and are a powerful means for studying, monitoring and managing landscapes of cultural relevance.

Landscapes

Natural Features, Formations and Sites
UNESCO (1972) uses the term 'Natural Heritage' for physical, biological, geological and physiographical features, formations and sites of outstanding value from an aesthetic or scientific point of view. Far more than 100 natural heritage sites are designated by UNESCO as World Heritage and many others are receiving various degrees of protection under state or local legislation. From a conservationist's point of view, it would be desirable to keep human interference completely away from these areas. Since, on the other hand, many visitors are attracted, management guidelines have to be prepared and enforced and a monitoring process is needed to detect and prevent unwanted changes. Nevertheless, natural features, formations and sites are subject to changes caused by nature itself (UNESCO Operational Guidelines specifically mention on-going geological processes and on-going ecological and biological processes) and it would be a misinterpretation of conservation to keep these natural processes away from the objects.

Cultural Heritage Objects and Landscape
Meanwhile, more than 600 objects of 'Cultural Heritage' are designated by UNESCO as World Heritage sites. This is just a choice selection of a heritage comprising innumerable single monuments, groups of buildings or historical and archaeological sites of outstanding value for historical, artistic or scientific reasons. Although cultural objects are man-made, UNESCO in its Convention mentions landscape in this context, too ("buildings because of their ... place in the landscape", "combined works of nature and man").

In fact, no cultural heritage object can be understood without taking the surrounding landscape into account. Human dwellings have to use places in the landscape where an optimal protection from natural forces (weather, flooding) and enemy attacks is possible, and the supply of essentials (food, water, kindling) is assured as well. Special topographic features are chosen in all religions as places of worship or as sites for the location of divine buildings. Sovereigns chose special places to erect their palaces and mausoleums and often changed the surrounding topography, too. Cities developed along trade roads or close to bridges and fords. And even when a small arrow tip is found somewhere in the fields it should be studied why it was lost here and not somewhere else in the landscape.

The location of all those objects in a landscape was a result of practical or metaphysical considerations and rules. Thus, it is not sufficient to examine and document the object itself. Far more often than presently done, the landscape surrounding a cultural heritage object should be considered, studied and documented, too. If the present topography is surveyed, mapped and visualized, historic evidence may be used to reconstruct landscape development from ancient to present times. At the same time, conservational hazards originating from the present topography (erosion, slides, flooding) or landuse (agriculture, industry, traffic) can be foreseen and possibly prevented.

Cultural Landscapes
There are cases where natural and cultural criteria of a landscape cannot be separated. The value of such landscapes, being both, of natural and cultural significance, was hard to define with the 1972 UNESCO Convention. This is why UNESCO revised it in 1992 and adopted three categories of ’Cultural Landscapes’ (Roessler, 2000):

Landscapes Intentionally Designed and Created by Man. This embraces garden and parkland landscapes constructed for aesthetic reasons. Often, but not always, buildings and ensembles are part of those landscapes.

Organically Evolved Landscapes originate from an initial imperative (social, economic, administrative, religious) and have developed their present form in an evolutionary process in close interdependence with the natural environment. This evolutionary process may have come to an end in the past (’relict’ or ’fossil landscape’) or it is still continuing (’continuing landscape’). Where a relict landscape needs conservational methods to preserve the site, a continuing landscape needs management plans and measures to allow evolution without destroying its outstanding value.

Associative Cultural Landscapes may show no man-made evidence at all (thus, from a materialistic point of view just being natural landscapes), but powerful religious, artistic or cultural associations of the natural element attach special importance to those landscapes.

Landscape Documentation

Virtual Landscapes
A landscape can only be experienced in all aspects when visited at its original location. Obviously, no virtual landscape can match reality. It should be noted, however, that not everybody is physically or financially able to visit any landscape of interest. Many landscapes can only be reached after a strenuous journey. Other places are located within areas of war or turmoil and - unfortunately - one has to be prepared that this may happen to places that have been safe so far, too. In these cases, a virtual trip through a digitally created landscape can be the only way to explore an area.

A person, moving on the surface of a landscape, can overlook only a very limited part of this landscape. Many impressive landscapes (e.g. a river winding through a mountain range) can only be observed in total from locations far above the ground. Even when such viewpoints can be reached by airborne vehicles, the vision will be limited by the vehicle’s structures as well as atmospheric effects. Virtual landscapes do not impose any limits to the observer in this respect because any observation point can be reached to have a look at the scene and virtual trips through and around the place at a course of one’s own choice can be undertaken. These are reasons why a virtual representation is of much more importance for the documentation of landscapes as compared to virtual images of other objects of cultural heritage.

Digital Elevation Models and Texture Overlays as Elements of Virtual Landscapes
A landscape documentation can just be an abstraction of the real world. No matter which method is used, only a selection of the landscape’s features can be surveyed, recorded and visualized. In order to describe the morphologic features of a landscape, elevations of single points are recorded. This data set, where spot elevations are attached to selected vertices are referred to as Digital Elevation Model (DEM). It may consist of single raster points arranged in a regular grid pattern, single points in an irregular pattern (e.g. from tacheometric measurements) or contour lines (from existing maps). Since elevations for points between these data have to be interpolated, there should be enough points in the DEM to describe the topographic surface adequately. If sharp edges are present, the DEM should include break line information to avoid smooth interpolation across these lines.

To give information about the features covering the topographic surface, a texture image has to be draped over the DEM. Raster type images photographed with digital cameras or scanned from maps or photographs can be used (Fig. 1). As long as this image is presented in vertical parallel projection, a simple map type 2D representation of the landscape is achieved. Shading may help to experience the third dimension; contour lines are more precise but not as easy to conceive. If all data are available in digital form, perspectives can give a much better impression of the landscape and its topography (Fig. 2). Animations with smoothly changing relative positions of camera and landscape result in a better three-dimensional perception and allow a good interpretation.


Fig. 1: Combination of a DEM and a map to produce a perspective view of a part of the Rhine River valley (Emmel 2000).

Surveying and Imaging Techniques
Digital Elevation Models can be derived by several different methods. For detailed DEMs, tacheometric surveys or stereophotogrammetric measurements from aerial photographs should be used. Larger areas can be surveyed using aerial photographs of smaller scale or even GPS measurements. Some Earth observation satellite sensors supply images suitable for stereoscopic vision by taking images from two different locations of an orbit, looking forward and backward towards the same ground location (in-track stereo) or by taking images from neighboring orbits (across-track stereo). These images from digital sensors or scanned photographs allow automatic DEM generation by matching techniques, thus supplying a very powerful and economic method to create DEMs for large areas.

Texture overlays are always needed in raster format. Nevertheless, vector information from maps, plans CAD or GIS systems can be used if scanned or converted to raster data (Fig. 1). Scanned aerial photographs or digital satellite images are ideally suited to generate overlays (Böhler et. al. 1997, 1999). Their potential has hardly been recognized anywhere in cultural heritage documentation. In the meantime, raster data with one meter pixel size are provided (SpaceImaging 2000). So, textures of high quality can be used (Figs. 2, 3 ), even when aerial photographs are not available.


Fig. 2: Perspective view showing a part of a Tang emperor’s mausoleum in Shaanxi, China. IKONOS data have been draped over a DEM derived from a local tacheometric survey. Vector data from archaeologic prospecting (findings, tumuli, location of an ancient wall) are superimposed.



Fig. 3: Comparison of different satellite image resolutions. (From Böhler, Heinz 1999, complemented with an IKONOS image). Upper row from left to right: IKONOS PAN, KVR-1000, SPOT PAN Lower row from left to right: IRS-1C LISS III, SPOT XS, LANDSAT

Accuracy of data can become a critical issue when different data sources have to be combined. All data have to be geo-referenced in the same coordinate system. Small scale maps, for example, show intended displacements when several symbols are very close together. Therefore, map information may not coincide with other information (see railway line at the left valley side in Fig. 1). If high resolution images are used, the information intended for combination has to be of corresponding accuracy. In the case of one meter IKONOS satellite images, we found out that the DEMs previously used with LANDSAT and SPOT images did not have sufficient quality.

Geographic Information Systems
Geographic Information Systems (GIS) are an ideal means to manage data of cultural landscapes or cultural heritage objects in landscapes (Heinz 1997). Quite a few World Heritage landscapes and objects are already documented (and managed) using GIS-techniques (Roessler 2000). Special guidelines for the use of GIS in cultural resources management were developed (UNESCO 1999). A GIS provides tools to store, manage, analyze and visualize spatially related data. If a GIS with a temporal dimension is chosen, the evolution of a cultural landscape can be modeled and studied. Also, future developments and planning alternatives and their consequences can be explored in the same way.

Mapping and Visualization Techniques
If information is available in digital form, any combination of data can be visualized in any views and scales necessary for the observer. Vector maps or orthophotos can be shown separately or as a combined product. If combined with the DEM, virtual views from any observation point on the surface or above can give an imagination of a present, past or future landscape (Fig. 2).

Not every relevant detail can be seen in images, however. Cultural heritage objects may be too small, covered by soil or vegetation, or even removed for conservation or museum presentation. To show those objects, 2D- or 3D-symbols can be generated and stored in separate overlays. If included in a multimedia or GIS application, these symbols can be switched on and off depending on thematic attributes or historical periods, and linked with further text or graphical information (Fig. 4).


Fig. 4: Screenshot from a multimedia project (Eifel, Germany). Archaeological information is linked to symbols in a virtual landscape.

Perspective views, based on images draped over DEMs, do not give a proper visualization of buildings, however, since standard DEMs, being only 2½ -dimensional, can not handle vertical surfaces and the images do not provide the necessary texture. Here, separate data have to be obtained. This can be a very large task, but the expenditure can be reduced when only relevant structures are shown and when their geometry is reduced to typical features. For example, for a virtual flight over a 40 km part of the Rhine River (which will be proposed as a World Heritage Cultural Landscape) we have produced simplified 3D appearances for many castles and ruins which can be placed in the landscape model. They can also be considered as 3D symbols, and we experiment with a virtual flight where those symbols will appear in a scale larger than the scale of the landscape model (Fig. 5).


Fig. 5: 3D symbol of a castle placed in virtual landscape. Symbol is based on the real appearance of the castle and shown in larger scale than topography. (Holtkamp 2000).

Finally, landscape changes, caused by nature or man-made, can be modeled if multitemporal information is available. Consecutive images or videos, showing the historical change, are very useful for understanding natural and cultural landscape development. Landuse changes can be shown using different overlays. If elevation changes took place (e.g. erosion or earthworks), the DEM itself has to be changed depending on time. Since appropriate tools for DEM modeling are not available in software products, own procedures were developed and applied for a video showing the geological (mainly volcanic) and cultural (landuse changes, heavy quarrying) changes of a landscape (Böhler et. al. 1999, 2000). The video is presented in the poster session of this workshop.

Landscape Management and Monitoring
Considering the size and complexity of a landscape, adequate tools have to be provided for its management. After UNESCO had introduced legal instruments to recognize and protect cultural landscapes of outstanding universal value, and several cultural landscapes were included in the World Heritage list, it was soon realized, that "the resources management, e.g. the management and protection of the cultural values for which these sites have been inscribed , has to be improved" (Roessler 2000). It is obvious that surveying and mapping procedures, based on sufficiently complete and updated databases, are needed to document and monitor the sites and to allow predictive modeling of future developments. The role of Geographic Information Systems is of special importance in this context (UNESCO 1999).


Fig. 6: Three frames from a video showing the development of settlements over 100 years (Eifel, Germany). Settlement boundaries were digitized from old maps and superimposed on a recent satellite image.

Especially in the case of ’continuing landscapes’, where a large number of authorities and private owners may have to co-ordinate their actions and political decisions have to be prepared and mediated, planning and modeling future developments is of utmost importance. Changes within the landscape or in its neighborhood may impair or even destroy its cultural value. This is why UNESCO demands ’reactive monitoring’ and periodic reporting to ensure that World Heritage Sites remain undestroyed and can be kept on the World Heritage List.

Conclusions
Documenting cultural landscapes or landscapes around objects of cultural heritage has often been neglected because the surveying and mapping expenditure was too high. With modern surveying and visualization techniques, landscape documentation and visualization has become affordable and rewarding. Especially when combined with geographic information systems, virtual landscapes can be a powerful means to describe a landscape and to monitor and manage its conservation and development.

References
  • Böhler, W., Heinz, G., 1999. Integration of High Resolution Images into Archaeological Documentation.

    ISPRS WG V/5, WG V/2 Joint Workshop, Thessaloniky. Proceedings, pp. 166-172.
  • Böhler, W., Heinz, G., Scherer, Y., 1997. Using Satellite Images for Archaeological Documentation. CIPA Symposium Goeteborg. ISPRS Proceedings, Volume XXXII, pp. 226-233.

  • Böhler, W., Scherer, Y., Siebold, M., 1999. Visualization of a Landscape Genesis. ISPRS Commission V/SIG Animation Workshop Onuma. Proceedings, pp. 9-14.

  • Böhler, W., Heinz, G., Scherer, Y., Siebold, M., 2001. Video Showing a Landscape Genesis. ISPRS Workshop Ayatthuya.

  • Emmel, V, 2000.Virtueller Flug über den Mittelrhein. Diploma thesis, FH Mainz, unpublished.
  • Heinz, G., 1997. Aufbau eines Geo-Informationssystems zur Dokumentation archäologischer Befunde. -Tagungsband,

    2. Geosystems Fachtagung. - Germering .
  • Holtkamp, B, 2000. Entwicklung dreidimensionaler Symbole für eine Virtual-Reality-Anwendung. Diploma thesis, FH Mainz, unpublished.

  • Roessler, M., World Heritage Cultural Landscapes. Identification, Conservation, Monitoring. ISPRS

    Proceedings, UNESCO-ICOMOS-CIPA-ISPRS World Heritage Session. Amsterdam 2000.
  • SpaceImaging, 2000: www.spaceimaging.com

  • UNESCO 1972. Convention Concerning the Protection of the World Cultural and Natural Heritage, revised 1992.

  • UNESCO 1999. GIS and cultural resources management. A manual for heritage managers.
© GISdevelopment.net. All rights reserved.