Combining Maps and More at Ur
Spotlight on matching maps, satellite imagery, and aerial photos
Observing connections in spatial data with Geographic Information Systems
I’ve been in Iraq for the past two weeks. Part of that time has been spent teaching Iraqi archaeologists from Mosul University some of the latest techniques in analysis of archaeological data, and I’ve been using sample data from my work on Ur to do this. I’m accomplishing multiple tasks in this way: demonstrating the use of Geographic Information Systems and analyzing the site through the notes and maps that Sir Leonard Woolley created. GIS takes spatial data and displays patterns as well as quantifies connections in that data. We’re not yet to the point in our grant where we will be tackling the spatial element in full, but I’ve made an exploratory foray into it to pave the way, to show the power of GIS, and to help teach others.
The first step is collecting the maps to be analyzed. I’ve been analyzing these for some time, and I’ve made several blog entries about them. One thing I’ve found in reanalyzing my work and gathering them for the GIS is that I make mistakes (not that this was shocking to me; let’s just hope I can catch them in continued reanalysis). I was wrong when I said that the grid on Woolley’s maps wasn’t placed there until the 50s or 60s. It was there in the final map drawn at the end of the season in 1931. But this is the earliest indication I can find and it seems to have been placed there by the architect putting all of the maps together into one larger one of the entire site, rather than the way we would do it today, by setting a grid even before excavation begins. The notes almost never reference this grid—almost. The final Ur season shows the only indication of reporting finds to square on the site grid, and then only in the widely spread area known as CLW, the city wall. There are references to smaller grids Woolley sometimes put within a large building, but the larger site grid is another matter. I’ve also found that my estimation of original excavation house numbers may be slightly off, especially as concerns House number 7 (one I thought to be securely located). As we dig deeper into old records, we often find our first indications were wrong. The way to progress is to own up to these errors and move on with the new, more accurate assessments.
The next step is geo-referencing the old maps and photos. Geo-referencing places a map on the globe by linking identifiable points on it to their UTM coordinates. I’ve used a QuickBird satellite image taken in 2010 to help me with this, with the additional assistance of Google Earth.
Overlay of pictures can be accomplished in programs like Photoshop, but true connection to global coordinates cannot. Furthermore, accurate scaling in the many directions to compensate for potential warping at the edge of old camera lenses or slight errors in map grid-lines would also be very difficult in such a program.
Once I have a base map that’s keyed to the globe as accurately as possible, I can then add many other maps, as well as aerial photos taken while excavation was underway and other spatial data from notes and publications. This assists me and all archaeologists and historians who want to know more about Ur and the way it was excavated in many ways—especially in locating the earliest trenches. Woolley didn’t map these in, but he often tells us that something was found, for example, in TTA or TTB. These are references to Trial Trench A and Trial Trench B, the first he excavated, and we have an aerial photo taken in 1922 that shows both.
But there’s so much more that can be done with GIS that it almost boggles the mind. I’ll make another post at the end of June about some of the other capabilities; since I’m in Iraq, it may have to be short. Unfortunately, I’m not at Ur, since it’s still politically difficult to get permission as an American to go to southern Iraq. I’m in the northeast, the Kurdish controlled area, but I hope to go to the Sulaimaniya Museum to see their objects from Ur at some point while here.