I whipped this map up for a reader researching his father earlier today and thought I would share. The image shows the layout of the internment Camp at Farnham, Quebec overlaid on some modern satellite imagery from Google.
Camp A, as it was initially known, opened in October 1940 and initially held civilian internees and refugees from the United Kingdom. The camp closed temporarily in January 1942 but reopened in April and was used to hold Enemy Merchant Seamen (EMS). In late 1942, the EMS were transferred to Sherbrooke and replaced by German combatant officers and a smaller number of Other-Ranks, serving as the officers’ servants and orderlies. It once again closed briefly in June 1943 before re-opening again as an officers’ camp in September 1944. The camp then remained open until June 1946.
Like most of Canada’s internment camps, the buildings were salvaged and torn down, the barbed-wire fences removed, and the guard towers dismantled. Today, the site is occupied by a Water Treatment Station, a gas station and Tim Hortons, and a Fire Station. For those interested, here is a link to the location.
You would be hard-pressed to find a PoW camp or labour project in Canada that did not have an attempted escape attempt or, in a few isolated cases, a successful escape. The labour project run by the Erie Peat Co. employing Enemy Merchant Seamen (EMS) near Port Colborne, Ontario was no exception.
Having opened in August 1943, the camp employed some fifty EMS from Camp 22 (Mimico, Ontario) in a peat-cutting operation in the Wainfleet Bog. Not all were in favour of having PoWs in the area, particularly with the Welland Canal nearby. Despite the EMS being non-combatants and the canal being under constant guard, many protested the presence of “the enemy” to a vital transportation route for the war effort. Among the first labour projects in the country, the success or failure of this operation had the potential to influence the future of PoW labour in Canada so it was of significant concern when five PoWs went missing two weeks after their arrival.
In the early morning of October 11, 1943, PoWs Gerard, Hoffmann, Kaehler, Krause, and Schluter left the camp, making their way east across the peat beds.
Gerard, Hoffman, Krause were picked up shortly after, transferred back to Camp 22, and sentenced to twenty-eight days detention. Kaehler and Schluter managed to avoid capture for another day, their escape coming to an end on October 13 in Windsor, when a civilian observed two “apparent foreigners” hesitant to enter a restaurant. Suspecting the two were up to trouble, he notified the RCMP who arrived shortly after and took the PoWs in custody. Both had ditched their PoW uniforms for civilian clothing.
Escapees were understandably hesitant to provide much detail about their escape and their time on the run so I have found that apart from the camp and the location of their capture, much of the detail in-between does not make it into the police reports. Fortunately for the RCMP (and me, seventy years later) Kaehler and Schluter were not only found with an Ontario road map but a list of the towns they passed through on their way to Windsor. Upon further questioning, the PoWs revealed they had hitchhiked along Highway 3 until they reached Windsor. Trying to reach Detroit, their route, as seen below, took them through Port Colbourne, Canborough, and Leamington, and finally Windsor.
The escape was eventually blamed on the poor security measures and an inadequate civilian guard force, prompting the Veterans’ Guard to take over security shortly thereafter. Despite constant security concerns, the project remained open until November 1945.
As they escapees did not provide their interrogators with their final destination or any contacts, the goal of their escape (apart from freedom) remains unknown – perhaps they were trying to make it back to Germany or maybe they wanted to disappear in the United States. Regardless, their route sheds some light on PoW attempts to escape, showing these men avoided the closest border crossing at Fort Erie and Buffalo, likely thinking that would be the first place the guards and RCMP would look.
This post is long overdue but better late than never!
About a year-and-a-half ago, I started using iGIS, an iOS app that allows you to visualize geospatial data on your mobile device. My Professor, Josh MacFadyen, had demonstrated the possibilities of this app in our digital history class and I was eager to find ways to apply it to my own research. I had fiddled around with the app outside of class – see my post here – but didn’t have a chance to test it in the field.
In late summer, I finally had a chance to put it to the test. Using maps and aerial photographs I had georeferenced and converted, I uploaded my files to both my iPhone and an iPad and set off. Now the PoW camp I was visiting is somewhat remote – the nearest cellular signal was ten kilometers away and, needless to say, there was no wi-fi. Not sure how the program would work just using the devices’ internal GPS, I pre-loaded the satellite imagery (from Google) – a handy feature when you are in the bush – which would let me compare the historic imagery and maps with what the site looks like today.
An hour by bike and I was at the site. I pulled out my phone and was pleasantly surprised to find that my phone’s internal GPS was accurate enough to show my location. As you can see from the image below, there are only a few signs hinting at the site’s history – all the more reason to find new technologies to fill in the gaps. The rectangular shape to the upper right of my position is the foundation of the camp’s powerhouse.
Things kept getting better for I discovered that the iPad worked as well – I was surprised for I wasn’t entirely sure that the iPad I was using had an internal GPS to begin with!
Having thrown in an aerial photograph from 1949 (above) and a forest inventory map from the 1930s (below) and I was in business! Adding a layer to show the physical layout of the camp as it appeared from 1943-1945 helped visualize the site and pinpoint some of the camp’s buildings and locations of interest.
While this is fantastic for my own research – I was able to identify tree stands cut by PoWs and see how the area has in some cases recovered or in others changed – it has important applications for public history as well. That same week, I led an interpretive PoW wagon tour, taking thirty people out to the site. As part of the tour, we provide visitors with a guidebook listing the buildings and explaining the history. Along with this we include a traditional map showing the layout of the site. While some of the footprints of the former buildings are visible if you look closely, it isn’t always easy to visualize the camp’s layout. However, with me leading a tour with an iPad, visitors were able to see exactly where they were standing in relation to the camp’s layout in the 1940s. Combining this with historic photos of the site and screenshots from my 3D model, visualizing the site became a whole lot easier!
That same week I visited the former PoW camp near Mafeking, Manitoba with the Beranek family, whose father/grandfather worked there in the latter war years. As I had only been to the site once before and was therefore much less familiar with this camp’s history than Riding Mountain, I relied pretty heavily on iGIS and an aerial photo to orient myself. I was also able to add the GPS waypoints I had taken on my handheld GPS the year before, seen as yellow dots in the image below.
Needles to say, using geospatial data on a mobile device opens up a realm of possibilities for both studying and sharing the past. While my use of the app and sources were fairly basic, I was able to visualize the site’s history, explore areas that I may have otherwise overlooked, and use non-traditional sources to better understand the relationship between history and the environment. Combining this with the ability to allow visitors to interact with landscape in new ways and let them explore on their own terms, I am eager to see how historians can use these technologies in the future.
With a major research trip (and some field work) planned for the summer, you can bet I’ll be bringing iGIS to help me explore and share the history of PoWs in Canada.
One of my more recent interests is the work of the Canadian Forestry Corps (CFC) during the First World War. The CFC was raised specifically to supply the Allied armies with a desperately-needed resource: timber. Whether it was to needed to support dugouts, hold barbed-wire, or aircraft frames, wood became one of the most basic resources of the First World War. Composed primarily of lumberjacks and foresters, the CFC eventually operated in France, England, and Scotland and succeeded in preventing a timber shortage. For a bit more of a background, you can read another of my posts here.
The CFC was organized into numbered companies of about 100-200 men and were assigned specific forests or woodlands to cut. One of these companies, the 107th (originally No. 7), received a timber limit in Keppernach, Scotland (South of Nairn in the Scottish Highlands). The area in which the company found itself in, the Forest of Cawdor, was no comparison to the Canadian “wilderness” they had left behind. Between 1825 and 1854, over fourteen million trees had been planted here, in an area covering just over 4,000 acres. The majority of the area was populated with Scotch Fir, the most-desired wood for wartime use.1
The company’s advance party of fifteen men arrived in late July 1916 and within ten days, had a Scotch mill up and running. A Canadian mill was built in August and the mills ran almost continuously until the camp closed in August 1917. By September 1917, the company had cut 6.5 million F.B.M. (foot board measure).2
As I’m interested in the relationship between forests and war as well as forest regeneration, I decided to see what I could learn using GIS. Fortunately, the online war diary of the 107th Company includes a relatively uncommon map of the company’s operation at Keppernach.
Even more fortunate, when you search Keppernach in Google Maps, the first (and only) result is Keppernach Farm and takes you to the exact location the 107th Company worked in (see below). This saved me quite some time as I would have otherwise had to search all over Nairnshire for roads that matched up with my map.
In trying to locate some contemporary maps of the site, I discovered that Scotland, and Great Britain for that matter, has some great HGIS sources available online, notably A Vision of Britain Through Time and Old Maps Online. While some of these require you to live in Britain (A Vision of Britain) or charge for digital copies, I was able to get some high quality and detailed maps (here’s an example) from the National Library of Scotland. These maps were produced by the Ordnance Survey and published in 1905 so they provide the perfect backdrop to CFC operations in the following decade.
I then threw everything I had into Quantum GIS (QGIS) and georeferenced the maps.
Then adding the CFC map (in red)…
While the hand-drawn CFC map doesn’t line up perfectly with the other layers, it still provides a sufficient sense of what the operation looked like. I was somewhat surprised to find that from the air, it is impossible (for me at least) to tell that this forest had been completed cleared in 1917 and that the roads, pathways, and borders remain virtually unchanged. I also found it interesting in that the forested area had actually expanded (likely the result of interwar afforestation efforts brought about by the wartime shortage of timber) and that you can still make out the site of the Canadian camp (large rectangle on the left).
For the moment, this is a chance for me to refresh my GIS skills (which I apparently lost while home for the holidays) and a quick experiment to see what I can extract from the archival records. While it is very much a microhistory, I’m hoping that it will lead to some new information and insight into the role of the Canadian Forestry Corps.
Stay tuned to see what else I can find!
1. Historical Record, Keppernach Camp War Diary – 107th Company, War diaries – No. 51 District, Canadian Forestry Corps, Reel T-10901-10902, Volume 5018, Series III-D-3, RG 9, Library and Archives Canada.↩
The following video is the result of a digital history assignment that I’m currently taken. The assignment tasked us with using digital methods to examine a significant or interesting landscape and naturally I chose the site of the Riding Mountain Park Labour Project in Manitoba’s Riding Mountain National Park. For those not familiar with it, this project employed 440 German PoWs in a woodcutting operation from 1943 to 1945 in an effort to prevent a predicted fuelwood shortage.
My first task was to find the sources. While I have a fairly sizeable collection of textual records relating to the camp’s history, maps and other spatial information are, for the most part, missing. Instead, I turned to aerial photographs to fill in my record gaps. Little did I realize how much I could learn from them!
With the assistance of the National Air Photo Library, the Manitoba Land Initiative, and a staff member at Riding Mountain National Park, I was able to assemble a range of coverage from 1931 to 2009. The next step was to import them into a GIS program and georeference them.
With the photos georeferenced, I was now able to add information from my records. As a map of the camp’s layout has not survived, my first step was to create an outline showing the buildings’ shapes and locations. Fortunately the building’s footprints, with some exceptions, were still fairly clear, even in my photographs from the 1970s.
The next major step was to look at landscape change. Like a map of the camp, a map showing the location of the woodcutting area has also not survived. Using aerial photographs from the 1940s and 1950s, I was able to plot the extent of the woodcutting operation, which, as I discovered to my surprise, was almost entirely confined to the northern shore of Whitewater Lake. By comparing these photographs with modern ortho-imagery, the regrowth of the spruce population is quite remarkable. The Parks Bureau specifically instructed that the PoWs leave spruce trees standing in hopes of regeneration. As you can see from the video, the spruce population has [spoiler alert!] done exactly that!
Anyways, I’ve talked enough so on to the video. While this isn’t going to win any Oscars and I am certainly not Morgan Freeman, I hope that this video demonstrates how GIS and other historical methods can be applied to studying history.
Thank you to all of the individuals who helped, especially Josh MacFadyen, who put up with all of my constant questions!
One last thing; if you are interested in learning more, Josh and I will be delivering talks on Historical GIS for GIS day on November 20. For more information, please click here.
Having completed Monday’s HGIS workshop with Don Lafreniere, I have to say that, after going in with experience in Quantum GIS and some ArcGIS, I found the workshop significantly less daunting than my previous time with Arc. Overall, I’m impressed with what ArcGIS has to offer and I’m looking at different ways to incorporate its features into my own research. Right now, I don’t have the spatial data to do all of the things I’d like to but perhaps one day.
In the meantime, I’ve been continuing my research with Quantum GIS, which doesn’t have the same features as ArcGIS but still provides me with most of the tools I need at the moment. Over the past few weeks, I’ve been working with QGIS to map the site of the Riding Mountain Park Labour Project. This project, located in Riding Mountain National Park, employed 440 German Prisoners of War from October 1943 to October 1945.
Using a topographical basemap from Natural Resources Canada, I began mapping my project. Fortunately for me, Manitoba Conservation’s Manitoba Land Initiative provides a wide range of geo-spatial information for the province at no charge! I reviewed their sources and downloaded various maps including topography, hydrography, road systems, land parcels, and vegetation maps. Now it was time to add my aerial photographs!
With the assistance of the National Air Photo Library and the very generous support from Riding Mountain National Park, I’ve collecting aerial photographs of the site. At this point, my collection includes images from almost every decade from the 1930s to the 1980s. While the range in quality and scale, they provide an interesting look at the changing landscape of the area surrounding Whitewater Lake. I discovered that georeferencing natural landscapes is significantly more challenging than their urban counterparts. While the earliest period I’m looking at is the 1930s, it is interesting to see what the features that do change (marshes, lakes, creeks, rivers, trails, etc) but to also see the things that generally stay the same. Thankfully there are enough of the latter for me to put these images on the map.
Now, the next step was to add the other historical sources to my map. In my research for my undergrad and Masters, I’ve assembled a fair bit of information regarding the Riding Mountain Park Labour Project but surprisingly little spatial data. In my collection of approximately 3,000-4,000 files relating to the camp, I have a grand total of two historical maps. Thankfully, both are pretty interesting and relevant to my research. Both maps were produced by the Parks Bureau when they were trying to find a location from the camp. The first showed the area affected by a large forest fire in the late 1930s while the second laid out the camp’s initial boundaries. After overlaying these maps, I was able to convert the information to vector layers that I can now manipulate.
I was also happy to find out that I could add my own GPS data. The black line you see on the map below is the route I took in early September of this year.
I was also able to add Google’s Satellite imagery though it appears that the plugin is not compatible with OS Mavericks so that image will have to wait! I’m still working on a few things at the moment so stay tuned in the next few weeks for more!
For this past Monday’s Digital History class, we took a field trip to London’s Victoria Park to better understand what Spatial History offers to both historians and the general public. As part of this, we looked at using the free iOS App iGIS.
With the ability to visualize GIS data, including georeferenced images, on a mobile device, iGIS allows users to bring their data into the field. In addition, using your device’s GPS, iGIS can also show and track your location relative to your data. Interested in applying to my own research, I decided to try make my own iGIS map with an aerial photograph of the PoW camp in Riding Mountain National Park from the late 1940s.
In order to have the image overlaid in the correct location, it must be georeferenced using GIS (for more information on this, visit the Geospatial Historian). With this already completed, the next step was to use the program MapTiler to convert the image into a series of tiles that are stitched together to provide one coherent image. For detailed instructions on how to tile your map and then tansfer it to your mobile device, iGIS provides detailed instructions here. One issue I did have was that in using the free version of MapTiler, my original 120MB .tif was too large. As I couldn’t find a way to simply select a section of my image or compress it, I cropped a section of it and re-georeferenced it (if anyone has a better solution, please let me know!). With the smaller file, I converted into a tiled format, zipped it, and put it on my iPhone. Apparently I did everything right as it seems to have worked!
As the camp was demolished in 1945 and 1946, one of the problems I face in my interpretive programs is how to get people to understand what the camp looked like in the 1940s. In the past, we have used map handouts and handheld GPS units but I’m intrigued at the possibilities that this app offers in helping visitors visualize the past!
This past week, in preparation for an upcoming assignment, I’ve been trying my hand at some Historical GIS. I spent the weekend going through the the GIS tutorials at the Geospatial Historian which, following up from my earlier post regarding mapping with Google Maps Engine, I highly recommend!
The program of choice is Quantum GIS (QGIS), which is an open-source and free program. The tutorials at the Geospatial Historian walk you through the basics of GIS and provide all of the files you need to produce the desired result. I should note that the tutorials were designed for the previous version of QGIS but I didn’t have any trouble following along with the newer version.
As I’m still waiting on the aerial photos for my assignment, I decided to fiddle around with QGIS so I’m more prepared when I do get my files. With the Western University Library’s digital coverage of London for selected years between 1922 and 1955, I decided to look for a London subject to test. Having visited CFB London (home of Wolseley Barracks) and the Royal Canadian Regiment Museum this summer at the 2013 OMMC Conference, I decided to take a look at how the base changed as a result of the Second World War.
The first step was to find something to use as a base map that I could then georeference the photos to. Using QGIS’s tool to add existing maps to my project, I used the “WMS-Toporama” map from Natural Resources Canada. Setting up my project to the appropriate coordinate system, I now had to georeference the pictures (assigning photographs with their geographic location).
The end result looks like this:
Looking at the layers individually, you can see the transformation of the site as Canada mobilized and demobilized for the Second World War. I left the road layer on as this is what I primarily used for georeferencing. You’ll notice it isn’t quite perfect but for what I needed, it does the job! Wolseley Barracks is the U-shaped building in the middle of the left hand side of the picture.
Now that I’ve figured out the georeferencing part, it is time to start looking at what else I can do with this program!
One of my projects this past summer was to go through the all of the records I’ve digitized from Library and Archives Canada and create a database of every PoW Camp (large, traditional camps) and Labour Project (smaller work camps) in Canada during the Second World War. As this is going to be useful (I hope!) for my dissertation and because no one, to my knowledge, has attempted to do so in the past, I wanted to create database that I both could consult as a reference and also map. Today, continuing on this week’s map kick, I decided to see what I could do with the basic tools at my disposal.
Following the Geospatial Historian’s excellent tutorials, I decided to tackle the first lesson on Google Maps Engine Lite. Expanding on Google Maps’ basic principles, Maps Engine Lite allows users to import data and geocode it (taking a place name or coordinates and putting it on a map) while also allowing you to draw vectors and polygraphs. For anyone interested in looking at some basic mapping techniques, I’d highly recommend the tutorial as it was well-detailed and very easy to follow!
Anyways, on to the mapping. My database was based primarily on bi-weekly and monthly statistical reports produced by the Department of Labour that kept track of PoWs employed on labour projects and various statistical reports on internment camps. I should also note that there were gaps in my records that I have yet to fill so it is not a completely accurate depiction of PoWs in Canada. From these records, I was able to obtain basic details such as a general location (usually just a town or city name), the employer, the amount of PoWs employed, the number of guards, the nature of work, etc. As I had already spent the summer entering this data into excel, all I had to do was import the data into Maps Engine Lite and let it work its magic.
Maps Engine Lite, however, is not without its limitations. As my location details were town names rather than coordinates, Maps Engine Lite relies on Google’s data to convert these place names to coordinates. With dealing with 70 year old records, there were quite a few towns or locations that simply don’t exist (and some were just railway stops to begin with). Thankfully, you can go into the data on Maps Engine Lite to tweak the locations to update place names. The other problem I encountered was the limited amount of data that I was able to enter. As Google only allows three layers with one hundred points in each layer, I had to split my approximately 270 points into three different files which wasn’t a big deal but still took some more time.
Now, with all of the data, I used the information I learned from the tutorial to change the colours of the individual points to differentiate internment camps (in red) from labour projects (in various colours depending on the nature of work associated with it). The end result, as seen below, offers me the opportunity to visualize my research in a way that has not been presented before and now I can start looking at the relationships between PoW internment camps and labour projects and their location.
As this was only part of the first lesson, I’m looking forward to incorporating this research into some more sophisticated software like GIS and seeing what I can learn from my data!