Another case study can show the integration of GIS and various types of remote sensing data for a rather unusual application, archaeology. This research project is a long-term analysis of the application of Geomatics to study the interaction between different cultures and the landscape over time in Burgundy, France. An American interdisciplinary team has conducted research in the Arroux River Valley region of Burgundy for over two decades.
The team uses an integrative approach termed historical ecology: the multiscalar analysis of the interaction of culture and the environment over time. A period of over 2,000 years (from the Celtic Iron Age, Gallo-Roman, Medieval, to the present) is being analyzed. The overall goal of the research has been to understand long-term interaction between these very different cultures and the physical environment. Various remote sensing, GIS, and GPS technologies have been used to support this work, which has been conducted by Dr. Scott Madry.
This map of France shows the location of the study area circled in red.
The area of Burgundy, whose chief city is Dijon, is shown in an extract from an Atlas map; note the location of Autun.
A feel for the landscape of parts of Burgundy is provided by this ground photo.
The study area is shown from space with this SPOT image that has been merged with DEM data to give a three-dimensional sense of relief (somewhat exaggerated); Autun is in the dark area in the upper right:
Along the heights above the river a series of Celtic hillforts are located. These have figured significantly in the history of the region, as well in the research activities. Foremost of these is Mt. Beuvray, in the northwest corner of the study area. This impressive mountain was the site of the great Celtic city of Bibracte, capital city of the Aedui, the powerful Iron Age Celtic polity whose territory was centered on the Arroux. Bibracte reportedly had a population of over 30,000 in 52 B.C., when Julius Caesar quartered his legions there. Other, smaller, hillforts follow the river to the south, while others follow the opposite side of the river.
The major modern city in the valley is Autun (see photo of the ancient Roman city walls as they look today). It was founded by Julius Caesar as Augustodunum Aedorum, home of the relocated Celtic inhabitants of Bibracte after the Roman conquest. Caesar's conquest of Gaul, including his time in this area, was documented in his Gallic war commentaries "De bello gallico commentarii".
While the region retains a mostly rural character, there are several modern threats to cultural resources in the region. The most serious of these are a series of gravel mines (sabliers) that run along the banks of the river. These mines are taking much of the land immediately adjacent to the river. The mines are destroying a large number of archaeological resources in the area, and one aspect of the project is to locate and document these sites before they are destroyed.
Archaeologists have been using aerial prospecting and photography since just after World War One to locate buried structure, roads, and other features of ancient landscapes. Several archaeologists who served as pilots in that war noticed strange circles and square patterns from the air that were not visible on the ground. After the war they returned to these locations to find that these were actually archaeological sites, ancient roads, etc. Early pioneers of 'aerial archaeology' conducted the first such surveys in Europe and the Middle East in the 1920's. Charles Lindbergh also conducted similar surveys in Central America in the 1930's. How is it possible to see ancient landscapes and the remains of structures that can be a thousand years old?
Faint lines and color changes visible from the air are often invisible on the ground, and can be caused by buried cultural remains. Aerial archaeologists refer to these as crop, soil, and shadow marks. Crop marks form because there can be notiecable variations in crop vigor, color, or height when crops or natural vegetation grows over buried walls or other cultural remains. These are called 'negative' crop marks, as the crop is less vigorous due to the lack of moisture or root vigor caused by the buried walls. The opposite are 'positive' crop marks, where the crops are taller or more vigorous when growing over pits or post holes. There is more moisture and better root growth, so the crops grow better. These differences are heightened in times of crop stress, such as a drought. In the major French drought of 1976 thousands of new archaeological sites were discovered from the air. Soil differences also can be visible, such as where a road or ditch was filled in with soil from a different place. These are called 'soil' marks. Small variations in topography causing shadows that are visible early or late in the day, these are called 'shadow' marks.
An excellent source of remote sensing information for archaeologists is archival aerial photography, the older the better. U.S. Army Air Corps aerial recconnaissance photos of the region dating from September 1944 were acquired from the U.S. Defense Intelligence Agency using a Freedom of Information Act Request (FOIA). These photos were acquired during the Allied push into the region near the end of the Second World War. Over 200 black and white vertical aerial photos of a scale of approximately 1:40,000 were acquired and have been manually and digitally analyzed to search for archaeological sites, roads, etc. Numerous features have been located and mapped.
This project has used a variety of modern remote sensing data. Since archaeological remains are very small, airborne data are very useful in conjunction with satellite data.
The Aries scanner system is a French airborne digital radiometer system that was used on this project. It was constructed by the Laboratoire de M�t�orologie Dynamique (CNRS, France). It has two channels, normally set with one in the area of the visible and near infrared, and the other in the thermal portion of the spectrum (Perisset and Tabbagh, 1981:185). Internal calibration of the thermal scanner can provide apparent temperature recording capability. The spatial resolution of the data is dependent on aircraft elevation, but 1-2 meter data are typical for missions such as ours. For this project, the ARIES scanner system was mounted in a single engine Pilatus aircraft, and three corridors within the research area were flown in 1987.
This airborne thermal scanner image shows the location of the Roman villa structure which has been destroyed by a gravel mining operation (shown by a star on the image). The field is now a lake. This is a fairly common situation in Europe, and part of the project activities is to locate archaeological resources so that they can be studied before they are destroyed by modern activites.
A somewhat more ambiguous example appears in the image below. This is the area in the upper left of the above infrared image. The "arrow(>)" marks point to a faint but perhaps meaningful curved (arc) pattern which may also be an ancient wall or some other buried structural feature.
The Autun project has used a variety of satellite imagery over the years, including Landsat MSS (80 meter) and French SPOT (20 and 10 meter) data. The SPOT data has a spatial resolution of 20 meters for multi-spectral data, which records information in three bands of the spectrum, and a 10 meter spatial resolution for a panchromatic band. The resolution of these images available from space can provide significant improvements in the utility of these data for regional archaeological and environmental applications, especially (as in France) where the field size is very small. Here is a false color SPOT image that includes part of the research area. In this image, Autun is at the upper right and Mt. Beauvrag at the upper left.
Accurate modern landcover maps were produced using SPOT satellite imagery. The upper part of this next image/classification corresponds to the above SPOT scene; look for the blue horseshoe pattern to fit it in.
Canadian RADARSAT-1 satellite imagery has been acquired for the region on 4 November, 1998, shown below. With a spatial resolution of 8 meters, this system is different from SPOT or Landsat in that it is an active radar system that sends its own burst of electromagnetic radiation down to the ground which bounces off the surface and recorded by the satellite. This system can operate day or night and through cloud cover. It provides a different and new way of visualizing the area. The first scene below shows the area around Autun. Below it is another Radarsat image that illustrates the hilly topography in parts of the study area.
The GIS data base covers an area of about 30 by 60 km, covering the majority of the Arroux River Valley and its immediate environs. The current basic raster layers of the GIS data base include: elevation (generated from the French digital elevation data), aspect (derived from the digital elevation data), slope (derived from the digital elevation data), SPOT images (20 meter false color infrared), SPOT images (10 meter panchromatic), land use/land cover maps (derived from Spot image data), geology (generated from 1:80,000 geology map of the upper 2/3 part of the region), faults (from the same 1:80,000 geology map), hydrology (from the three 1:50,000 topo maps), modern roads (from the three 1:50,000 topo maps), ancient roads (from project information and old maps), known Celtic hillforts (from project information and old maps) data layers showing different distance categories, or buffer zones, from roads, streams, faults, archaeological sites, hillforts, and ancient roads (also generated from the data above). Additional data have recently been added that were derived from the 1:25,000 maps.
Two of the thematic maps listed above are of special utility in the GIS analysis described below. On the left is a map of the Celtic road network developed in or after Roman times. On the right is a map of known hill forts of the time. These are helpful in calculating distance from any vantage point to specific features.
An original year 1659 map of the region and two 1759 maps of the research area were digitized using 4,000 by 4,000 digital CCD array device that converts maps or photographs into digital format. The 1659 map which was scanned from the original, entered into the GIS, georeferenced, and entered into the GIS system as a data layer like any other. This map is quite general in its spatial accuracy, but it is the oldest map yet found that covers the research area. Here is part of this map:
Two 1759 maps were also scanned into the system and patched together to make a single data layer. Here is the Cassini map centered around Mt. Dardon. It clearly shows the roads, rivers, topography, and every individual house and structure.
These 1759 maps were produced for the famous Cassini triangulation survey of France conducted in the mid 1700's. This massive work took three generations of the Cassini family to complete, and was the first accurate survey of an entire nation using modern surveying and mapping techniques. Reproductions of the maps are available for all of France from the Institut Geographique National in Paris. These historic maps are extremely detailed and accurate, much more so than the older map we possess created 100 years before it.
The Cassini maps provide us with the oldest, reasonably accurate, record of the location of roads, bridges, towns, and villages of the area. It also shows information about the vegetation and land-cover of the area at that time. The maps contain an astounding level of detail, and have been digitized to represent a series of GIS layers representing the cultural and environmental makeup of the area at that time.
One of the more interesting results from the GIS analysis was the line-of-site analysis (Madry and Rakos 1996). This GIS technique allows one to determine what parts of the landscape are visible from any given location. The research demonstrated that the old Celtic road network connecting the hillforts of the area tended to follow within the line-of-sight of the hillforts, rather than take more direct paths (as originally proposed in Madry and Crumley 1990). We have run the same line-of-site analysis from four locations on each of the known Celtic hillforts in the research area . These four line-of-site maps for each hillfort were then combined to generate a map of complete inter-visibility from each entire hillfort, assuming that the forts were manned by watchers from each of the four "corners" of the ramparts. An eye height of 5 meters above the terrain was used, assuming that towers of just over 3 meters height were strategically located around the ramparts (and that the eye level was just under 2 meters above that height).
These individual line-of-site maps are combined to produce the map below which shows (in red) the total portion of the research area and each ground transect that is within line-of-site of each hillfort, and for the total network of hillforts in the region. This analysis shows that the Celtic roads definitely tend to follow paths that stay within the view of the hillforts, even if they are less direct or require a steeper climb. Additional work was done to model the location of Celtic and Gallo-Roman roads where the exact location of segments is not known.
Site location modeling is a useful GIS product for archaeologists who wish to locate and protect unknown cultural resources. It allows us to model where archaeological sites of a given period may be located, based on the known site locations and various environmental and cultural data in the GIS. Predictive models were developed on this project using site data generated from field surveys conducted in 1978 and 1979. A model based on environmental and cultural data was created that accounts for 78.9% (45 of 57) of all Gallo-Roman sites in only 29.2% of the total area that was field surveyed. The same model also includes 69.2% (36 of 52) of the Iron Age sites and 80.3% (49 of 61) of the Medieval period sites located in the field survey. This model was then generalized to include a much larger area surrounding the transects (4.7 times as large). New layers in the GIS containing these locations were produced along with new maps showing the areas with the highest probability of site locations. These areas of higher probability of archaeological sites have a high correlation with areas that are threatened by current gravel mining activities in the area.
Current research involves analysis of aerial photography and site surveys using GPS in the areas predicted to have higher site potential by the predictive models. Coordinates of high probabililty areas are entered into GPS units, which are used in the field to search for the exact location of the archaeological sites. GPS receivers are also used in the aircraft while conducting aerial prospecting and photography. The GPS allows us to locate our position and the location of the photographs much easier than looking on maps for individual fields while the aircraft is banking and turning.
These Roman sites tend to cluster along the rivers in flat bottomlands, and also along the Roman roads. Celtic living sites tend to be more in uplands and near the Celtic hillforts on the hilltops. The largest single area of high probability is the area shown below, next to the river, where the Roman villa was located and later destroyed. The gravel operations continue to work in the area, and researchers are trying to locate additional sites using these techniques before they are destroyed.
This project has used a variety of visualization tools to assist in better understanding both the data and the region. New technologies allow us to view our data in interesting ways from the laboratory when it is not possible to be in the field. For example, we can "fly" through the study area and look at different GIS layers draped over the terrain. First we need a digital elevation model (DEM) showing the topography of the region. For this project this was generated from the French 1:25,000 scale topographic maps.
Shown on the left above is the gray-scale DEM image of the Arroux river valley. Lower elevations along the river are shown in black, shading to white which represents the highest terrain in the region, including the site of the hillforts. From this a slope steepness map, seen on the right, can be derived using colors to indicate small intervals of slope values in degrees. These data can be displayed in several ways: one is the SPOT image of the study area with topographic effects superimposed, as we saw near the top of this page.
Drawing upon an appropriate topographic rendition, we can create perspective views that ultimately can be combined to produce a 3-D fly-through of the valley (described below), which shows the region in three dimensions as we fly up the river. We may choose to overlay any individual or combination of GIS data layers in these views.
As seen from an oblique aerial perspective, the GIS scene for an instant might look like this:
To generate the fly-by, we have applied a program that uses input data such as we have created to generate a movie-like run through the scene. Quicktime Virtual Reality (QTVR) lets us pan and zoom interactively on our computer (or over the web) to see the actual view from given locations in the study area. This is very useful in determining the accuracy of Line of Sight GIS computations, and also in allowing people to experience a virtual visit to the research area on the Web. To view the Quicktime VR movie you will need to download a free plug-in available from Apple Computer (the software runs in both Macintosh and Windows). For futher information, click on the Visualization word found on the Home Page of the Web site set up by this research team. That site also has background that supplements the overview on this Tutorial page and has the .mov file. It is your choice as to trying to download that file and run the movie. Unless you already have Quicktime, that program takes up to an hour to download.
The visualization is outward from the perspective centered around the summit of the Celtic hillfort of Mt. Dardon you have seen in the images. You can see the commanding view this hillfort had of the surrounding terrain, and portions of the Celtic ramparts which have survived intact.
Aerial imagery, and even more, GIS data are somewhat difficult for many people to relate to. They are abstract and it is hard for many people to relate what they see on the computer screen with what is actually on the ground. Visualization tools like QTVR allow us to bridge the gap between the imagery and reality, providing the ability for us to 'visit' the research area and compare the actual landscape with what we see on the various images and GIS layers.
This project has used a variety of tools to study the regional archaeological patterns of a large area in France. Particular emphasis has been placed on finding archaeological sites and then predicting the location of unknown sites before they can be destroyed by modern land use practices. It is the combination of new technologies and techniques, such as aerial photography, GIS, GPS, remote sensing, and visualization, that provides the greatest value to the researchers. Work on this project is continuing.