First World War Battlefields Map

How I Actually Build a First World War Battlefields Map

The biggest mistake people make is assuming these maps are just digitized old photos. They aren't. A usable First World War Battlefields Map requires stacking at least three separate data sources, each with its own coordinate problems, and then wrestling them into alignment. Most online maps you find are someone's quick Google Earth overlay that looks convincing until you zoom in near Passchendaele and realize the trench lines are shifted by about 400 meters because the original survey was done on a broken theodolite in mud. Start with the British Great War Atlas from the University of Toronto's Spatial History Lab. They've already done the hardest part, which is georeferencing the original trench maps. Then layer in the Imperial War Museum's collection of aerial photographs. These give you the ground truth for what actually existed versus what was drawn on official maps. The German equivalent is in the Bundesarchiv in Freiburg, though getting access to their higher-resolution sheets requires filling out paperwork that will test your patience. For anyone doing this work, the First World War Battlefields Map isn't a single product, it's a workflow. The actual deliverable is whatever comes out after you've spent weeks deciding which source to trust when they contradict each other. And they always contradict each other. A British artillery position shown on a January 1917 map might have been moved by February because the Germans spotted it and started shooting there.

The Technical Process

Here is what the work actually looks like day to day. You load a set of scanned trench maps into QGIS, pick control points that haven't changed in a hundred years like railway junctions or church towers, and georeference the raster. This is where most people burn three days because the original mapmakers used different datums. The British used the Clarke 1858 spheroid. The Germans used Bessel. If you don't transform them properly, your layers will drift apart by several hundred meters across the whole map, and nobody will notice until you try to overlay casualty data on top of it. Once the base layer is locked in, you add the aerial photos. These are the ones that matter most because they show the actual terrain as it was after years of shelling. A map might show a intact village, but the photo from 1917 shows it's gone. You use those photos to correct the mapped features. This step takes longer than anything else. I spent about two weeks just trying to get the Somme sector right because the photographs from different dates showed completely different drainage patterns, and the original maps didn't reflect any of that.

The Problem Nobody Warns You About

Here is the specific edge case that nearly made me quit this project back in 2021. I was working on a detailed map of the Third Ypres sector, around Polygon Wood and the Menin Road. Everything looked solid until I tried to overlay the German order of battle data against the British advance maps. The British maps showed a capture date of April 9, 1917 for a particular stretch of trench. The German combat logs from that same unit showed they hadn't been ordered to abandon that position until April 13. Four days of disagreement on what was actually held. The workaround was to stop trying to pick one source and instead build a confidence layer. I marked each trench segment with a colored border indicating which sources agreed and which didn't. Red for disputed, green for corroborated, yellow for single-source only. It took an extra evening of work but it saved the project from being misleading. A map that pretends certainty where there isn't any is worse than no map at all.

Software Choices

QGIS handles the heavy lifting for free if you know how to use it properly. The ArcGIS crowd will tell you otherwise but the georeferencing tools in QGIS are sufficient and the processing speed is better on older hardware. For the actual mapping and styling, I use a combination of QGIS for the structural work and then export to Inkscape for final labeling. Cartographic labeling in WWI territory is brutal because every field ambulance, dugout, and support trench has a name on the original maps and you need to decide what to include. There is a threshold where adding more detail makes the map unusable. Past a certain zoom level, you are just placing text labels on top of other text labels. I learned this the hard way on a 1:10,000 scale attempt at Vimy Ridge where the final product was basically illegible. Dropping to 1:25,000 and selecting features by operational significance rather than listing everything made the difference between a useful reference and visual noise.

Download and Resources

If you want to start without building from scratch, the Imperial War Museum's map collection at iwm.org.uk has a search function that lets you filter by sector and date. The Australian War Memorial also has a decent archive with georeferenced maps for the western front. For anyone who wants to collaborate, there is a community around the First World War Battlefields Map projects on GitHub where people share QGIS project files and the transformation parameters they found work. Joining that channel early will save you weeks of reinventing datum adjustments. Be honest about the limits. For the eastern front, the data quality drops significantly. Russian surveys from that period are spotty and the geographic references are often approximate at best. The Ottoman front is even worse. If you are mapping Gallipoli or the Mesopotamian campaigns, expect to spend twice as long on half the accuracy. You are better off relying on secondary scholarly atlases than trying to build your own from primary sources in those regions. The underwater dimension is also something most people ignore. Mine craters, underwater trench systems, and flooded zones are almost never shown on standard battlefield maps but they shaped the fighting. The Somme river crossings changed course multiple times during the war due to artillery damage. A static map of the river won't help you understand why a particular crossing point was contested in July versus October. That requires time-series mapping, which is possible in QGIS with the TimeManager plugin but adds another layer of complexity that most beginners skip over until it costs them.

The realistic output from a proper project like this, if you're working a single sector with reasonable source material, is about 40 to 60 hours of actual labor. Not including the research phase. If someone is offering you a finished map of the entire western front for free, it is either a simplified overview or it has undetected registration errors. There is no shortcut through the source verification work.