Visualising the last glacial maximum correctly
A few days ago, Swiss Geoportal (/the Swiss federal geodata infrastructure) tweeted a link to a visualisation of the largest ice extent over Switzerland – i.e. the last glacial maximum, LGM for short:
They used a rather famous map of the LGM and draped it over an elevation service as a showcase for the newly-developed (and charmingly accurately named!) 2.5D mode of their portal.
I remember seeing the LGM map during my studies: I’ve always been quite fascinated by it as the depicted ice masses are beyond comprehension and it has substantial explanatory power for current landforms.
Hence, I found it a bit of a shame to see the LGM map draped over the current DEM. This gives a very wrong impression of the actual topography 24,000 years ago. In my tweet I also hinted that all ingredients (i.e. iso- or contour lines of the ice surface) for building an accurate DEM of the ice surface are present in the map. Quite to my surprise I learned that vector data of the LGM map is actually available, but Yvo Weidmann beat me to building a DEM from the vector file.
I dabbled a bit in ANUDEM, but finally opted to use Yvo’s great work as a stepping stone to build my own
3D, errh, 2.5D model of the LGM landscape. I incorporated another DEM for the non-ice-covered areas and smoothed the transition between the ice and land surfaces. What follows are some comparison screenshots between Swisstopo’s and my version. Note how the ice bodies in my visualisation fill out large parts of the valleys, whereas they follow the typical u-shape of glacial valleys in the Swisstopo version.
You can access my interactive regional 2.5D models by clicking on the links below, or a low-resolution model covering all Switzerland by clicking here (allow time to load and please disregard the artefacts at the DEM boundaries).
The interactive visualisations have been built using QGIS and the formidable qgis2threejs.