Our immediate goal is to pre-process (i.e. clean up) declassified WWI-era maps of trenches in Northern France and Belgium, preparing them for statistical analysis. The down-stream analytical goal is to better understand the long-term legacies of warfare for local communities. Proximity to WWI trench warfare had profound implications for local political dynamics, economic recovery, and ecological health. Communities near the front lines endured significant physical devastation but also became focal points for long-term political and cultural transformations. The declassified military maps in this collection will allow us to precisely measure each locality’s exposure and proximity to sustained trench warfare in WWI, and to estimate its long-term effects on voting, economic activity, agricultural productivity, and other outcomes in subsequent decades.
Example of WWI trench data, Aubigny
These are GIS data in raster format (.tif
). Rasters are grid-based structures of equally sized cells (pixels), each representing a geographic location with a value for continuous or categorical attributes. In our case, the cells represent 4 \(\times\) 4 meter plots of land (resolution of 4m), and their values represent the presence or absence of trenches or other fortifications in WWI. The cells take a non-zero value if trenches were present (usually a 2 or 3, depending on whether the positions belong to allies or Germans), and are empty (NA, missing value) elsewhere. The above image shows an example of what these data look like.
The main problem we need to address right now is that some of the raster data are contaminated with “false positives”: raster values that falsely indicate the presence of trenches where there were none. The most common of these are street names, hill numbers, and other annotations that a geoprocessing algorithm might mistake for trenchlines, likely because these features were drawn in a similar color on the original map. If these errors remain, we will systematically over-estimate exposure to trenches in some locations, potentially biasing statistical analyses. The image below shows an example of these false positives: the black cells include street names and other info, not just the physical location of trenches.
Example of WWI trench data with “false positives” in need of removal, Amiens
Your goal is to remove as many of these “false positives” as possible. The instructions below tell you how to do this in QGIS. QGIS is a free, open-source, cross-platform geographic information system (GIS) software for creating, editing, visualizing, analyzing, and publishing geospatial data. Get it at qgis.org.
Project > Save As
in the main menu..qgz
(default) or .qgs
format.Project > Save
or using the keyboard shortcut:
Ctrl+S
Cmd+S
Plugins > Manage and Install Plugins
.Plugins > Serval > Show Serval Plugin
to enable the toolbar.Layer > Add Layer > Add Raster Layer
or click the Add Raster Layer button (a small icon resembling a grid with a plus sign) on the Data Source Manager toolbar....
) next to the “Source” field, navigate to your raster file (e.g., .tif
), and select it.
Dropbox
folder Dropbox/YZRA/Data/WWI/Raw
. They are all named something like reres_PARIS.tif
~/Dropbox
(Mac and Linux), C:\Users\YourUsername\Dropbox
(Windows).Add
to load the raster into your project. The raster will now appear in the Layers Panel and on the map canvas.Export > Save As
....
) next to the “File name” field to open a file browser. This allows you to specify the exact path and file name for your working copy.Working
directory in the WWI
folder (Dropbox/YZRA/Data/WWI/Working
) and enter a descriptive name for your working copy (e.g., copy_PARIS.tif
)./tmp/
or similar paths), as these files may be deleted when QGIS is closed..tif
file extension) and maintain consistent projection settings with your original raster.OK
to save your working copy.Properties
.OK
to apply changes and close the dialog.Polygon Selection Tool
Drawing Polygon
Apply NoData Button
NoData Applied
Ctrl+Z
(Windows/Linux) or Cmd+Z
(macOS) and Edit > Undo are not available while working within Serval.Export > Save As
.clean_PARIS.tif
) and location (use the Processed
folder in YZRA/Data/WWI/Processed
to store all completed files).raster_tracker.csv
in the YZRA/Data/WWI
folder. When you begin working on a file (and have created a working copy), change the value in the working
column for that file from N
to Y
. Save and close. Then do the same for the processing
column when you’re done.To Do
to Doing
and Done
as you go.Processed
folder without having to do any editing at all. But before you do that, be sure to closely inspect every inch of the map to make sure it’s really clean!Cmd
while scrolling if needed.Ctrl + +
(Windows/Linux) or Cmd + +
(macOS) for zooming in.Ctrl + –
(Windows/Linux) or Cmd + –
(macOS) for zooming out.Ctrl+S
/Cmd+S
) after major changes.Plugins > Manage Plugins
.Settings > User Profiles > New Profile
.