Some days I have moments of ah ha with my job.
Other days I have plenty of self-inflicted D’oh moments!
Last week I had several moments of both ah ha and D’oh in the same afternoon. In my last post, I told you about how the Provincial Archaeology Office tried to relocate some of the older sites in the inventory the old-fashioned way; by going out into the field and finding them. This week I am going to share with you a similar post, except this time we are going to use some newer technology to relocate some sites.
In 2011, I wrote two blog posts on the Intermediate Period in North West River and Sheshatshiu. The first post focused on Dr. William Fitzhugh’s work in NWR. Fitzhugh spent two field seasons, 1968 & 1969, working in the Hamilton Inlet area of Labrador. For those of you who don’t know Dr. William Fitzhugh is the Director of The Arctic Studies Center at the Smithsonian National Museum of Natural History. He used this information to write his PhD thesis: Environmental Archeology and Cultural Systems in Hamilton Inlet, Labrador: A Survey of the Central Labrador Coast from 3000 B.C. to the Present. In his thesis, Fitzhugh provided this map showing the location of his sites:
The Provincial Archaeology Office did not exist in 1968-69. Therefore, Fitzhugh’s site location information was not transferred to the maps used to keep track of site locations. In the ensuing years since 1968-69, many things have changed in NWR to the point where it is now hard to rectify Fitzhugh’s site locations with what the community looks like today. Because of not having transferred Fitzhugh’s site data, we could not confidently plot his sites. We knew they were in NWR but we could not say accurately where they were. Essentially the sites were displayed in our ArcGIS system as something like two starbursts; essentially, we had two sets of coordinates for more than 35 sites.
When I wrote the posts on the Intermediate Period in 2011, I tried to georectify Fitzhugh’s map with the PAO topographic map in ArcGIS. Georectify means to align digitally a satellite or aerial image with a map of the same area. In georectification, the images are layered on top of one another and a number of corresponding control points, such as street intersections, are marked on both the image and the map. These locations become reference points in the subsequent processing of the image. The GIS software then warps the images so they fit together and overlap exactly. When I tried to georectify the above maps I was not able to get enough control points so I couldn’t get the images to line up properly and therefore I could not accurately plot the sites.
As you can see, particularly from the shoreline in the upper left corner, the maps did not line up. Part of my problem was that I could not get enough control points and I used one control point that was actually way off. There is a small wharf just above the letter ‘H’ in NORTH on Fitzhugh’s map. I assumed incorrectly that this wharf lined up with the wharf on the colour PAO map that is just above the letter ‘N’ in North.
I recently had reason to go back to trying this georectification exercise for Fitzhugh’s NWR sites. That’s when I had both a moment of ah ha and a moment of D’oh! It occurred to me that I could retry this georectification exercise or I could just email Dr. Fitzhugh and ask him if he had a decent map or perhaps even coordinates for his sites. My ah ha moment! Please see the kid in the red shirt above. My next thought was, well, why had no one thought of that before now? My D’oh moment, please see Homer.
Within an hour or two, Dr. Fitzhugh responded to my email, unfortunately, he was not able to help with a better map or coordinates. Archaeology in 1968-69 was conducted very differently than it is today. While he couldn’t supply any information, something he said in his email was very helpful; he stated that he “…used the air photo to produce the map…” Looking at his map, the air photo was inset at the bottom left corner. Well, of course, the map was based on the air photo, why hadn’t I realized that before? A Homer moment, D’oh! Looking through our collection of air photos, I realized we had one that dated to around the same time. Looking closely at the air photo I realized that the wharf that I had been using as a control point on our colour map was not the same as the one on the Fitzhugh map. I scanned the air photo and enlarged it. (Unlike all the CSI episodes I have ever seen, the photo became grainier as I enlarged it, not clearer!) However, I was able to make out details on the air photo that I could line up on Fitzhugh’s map such as clearings and roads/paths. Then I saw the wharf in the air photo that Fitzhugh had drawn on his map and realized just how far off I was. So I loaded both into ArcGIS, selected my control points and they aligned very nicely.
Then I tried to line up the air photo with the colour map. I had no problem with that because I was able to line up the wharf in the correct spot and I was able to line up some roads/paths and the shoreline. Then I had an ah ha moment, if I can do that then I can line up Fitzhugh’s map with the air photo!
Now I was finally able to line up Fitzhugh’s map with the PAO colour map and thereby more accurately plot his sites!
The sites now seem to be spread out on the land as opposed to how they are dispersed on the first colour PAO map in this post, two starbursts. As an added bonus, I was able to test the accuracy of this new plot by looking at the location of FjCa-29 (the most northerly of Fitzhugh’s sites, near the middle of the map). That site is known as Graveyard Site, it would have been located somewhere within the graveyard that is the white square just below the dot for FjCa-29. Based on the new plot, the current distance of the site from the graveyard is just 20 metres. While this is not as accurate as I would like it to be, it is much better than the old plot.
The benefit of all this is that since we have a better idea of where the sites are we can better protect them and better predict where unknown sites maybe, while allowing development to occur.