The ‘Invisible Dead’ in Space. A post by Miriam Rothenberg (MA Archaeology, Durham University)

One of the many types of data being compiled by the ‘Invisible’ Dead Project is spatial data, i.e., where burials are located. I was tasked with georeferencing maps from Kathleen Kenyon’s excavations of Jericho in the 1950s to modern satellite imagery of the area. The excavation maps will help researchers from the ‘Invisible Dead’ Project figure out exactly where the burials they are studying are located. It will also mean that it is possible to make more accurate calculations regarding the total area excavated versus potential area occupied/used for burial.

Excavation Areas recorded by Kenyon (1960; 1965) digitized from scanned, georeferenced plans and displayed on top of a 15 m resolution Landsat Satellite Image (Band 8). Using modern imagery it is possible to identify features such as the edge of the settlement mound (Tell es-Sultan) and modern roads which can help to with georeferencing the image.

Excavation Areas recorded by Kenyon (1960; 1965) digitized from scanned, georeferenced plans and displayed on top of a 15 m resolution (Band 8) Landsat Satellite Image (Image courtesy of the U.S. Geological Survey).  Using modern imagery it is possible to identify features such as the edge of the settlement mound (Tell es-Sultan) and modern roads which can help us in georeferencing the image.

 The main difficulties with georeferencing these maps are that they were designed for a specific but limited purpose and are over half a century old. The maps were not intended to be manipulated in an electronic Geographic Information System (GIS), the technology for which didn’t exist when they were created. Some basic problems emerge from the fact that they were all measured and drawn by hand; they are often orientated well away from true north and are flat images that do not account for the curvature of the Earth. Perhaps the greatest challenge in working with them is that they were meant to be read in relation to landmarks that often no longer exist. It is not surprising that buildings, roads, and even whole neighbourhoods have been built, moved, or disappeared over 60 years, but this makes aligning old maps and new satellite images very difficult.

The first step in georeferencing non-topographic maps/excavation plans like these is to come up with an initial visual anchor point for the map’s general location. For these images, the key to figuring this out was Tell es-Sultan, the large settlement mound also excavated by Kenyon and located in the southwest corner of the study area. While the infrastructure has changed greatly over the last 60 years, the edges of the tell have remained largely the same.

I had four types of data to work with: a drawn plan of the total studied area, an old aerial photograph of the area, plans of smaller ‘tomb areas’, fitting inside the larger plan and the modern satellite imagery. The first task was to georeference the old photograph to the modern imagery. With this done I then, georeferenced the overall plan to the modern image, with help from the old photo. Finally, I could place the plans of the individual tomb areas onto the larger, georeferenced plan by matching their corners to the corresponding rectangles given on the larger plan and making sure they were orientated correctly.

Once I figured out generally where the image was supposed to fit onto the satellite map, the actual georeferencing was not that difficult. It felt a bit like a game, where the goal was to get the excavation map to line up as closely as possible with the modern imagery, while creating the least distortion. To do this, I matched known points on the map, like houses or street intersections, with their corresponding locations on the satellite photograph. The trick was that the more points I used, the more accurate the georeferencing became, but also the more likely the image was to become distorted. The computer keeps track of the distortion by giving a number, or ‘residual’ for each point. The game was to make the best visual match between the two images, while keeping the residuals as close to 0 as possible.

Area A Tombs. This example shows how, using GIS we can start to digitize individual tomb groups and further analyse the distribution of tombs dating to specific periods or containing certain objects etc.

Area A Tombs. This example shows how we can start to digitize individual tomb groups which will allow us to further analyse the distribution of tombs dating to specific periods or containing certain objects.

The georeferenced images are not perfect, but they are a start in mapping out where the Jericho burials were found and will allow us to consider the spatial relationships between tomb areas and their placement within the wider landscape. Some of the burial areas are missing maps entirely. Others have maps that, when georeferenced, are not at the scale that they claim to be, or have north arrows that do not point north, leading to some confusion. Still others are missing features or topography, or have these things in the wrong places. One of the problems of working with legacy data is that we cannot turn to the people who created them and ask for clarification or additional materials. We have to work with what we have- holes, inaccuracies, analog format and all. I hope to be able to refine the georeferencing of these images in the coming months, if necessary, to provide a more accurate spatial dataset. Once this is complete, the next step will be to digitize the locations of the individual burials within the studied area so that this locational data can be added to the database of information on the sites to make it more comprehensive.

References and Acknowledgements:

Kenyon, Kathleen M. (1960). Excavations at Jericho, Volume One.  The Tombs excavated in 1952-4. Jerusalem: British School of Archaeology in Jerusalem.

Kenyon, Kathleen M. (1965). Excavations at Jericho, Volume Two.  The Tombs excavated in 1955-8. Jerusalem: British School of Archaeology in Jerusalem.

Landsat Imagery is available from the U.S. Geological Survey.

Advertisements

2 responses to “The ‘Invisible Dead’ in Space. A post by Miriam Rothenberg (MA Archaeology, Durham University)

  1. Once the maps are digitized, is the spatial data itself going to be released? It seems like you’re doing a very great service for archaeologists- upgrading old maps- and it would be wonderful if the original files were available!

    • Hi Katy,
      Thanks very much for the message! The digitizing we are doing is fairly patchy at the moment, just as and when we need it and there are copyright issues with making some of the original map scans public. At the moment the project is just moving into the data analysis stage; however, it is our intention to eventually make the data from the project available for other researchers, we just have a fair way of analysis to go before we do this. So, keep watching this space and thanks again for the feedback!!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s