Doggerland — Visualising a Lost World
The project is open for collaboration with archaeologists or palaeoenvironmental scientists who want to use the reconstructed datasets. All processing scripts are published on GitHub under an MIT licence.
Doggerland was a vast, resource-rich landmass in what is now the southern North Sea. Inhabited by Mesolithic hunter-gatherers, it was gradually inundated as global sea levels rose following the last glacial maximum. Around 6,200 BC, a catastrophic tsunami triggered by the Storegga Submarine Landslide off Norway may have accelerated its final disappearance.
This project reconstructs that world at four temporal snapshots using GEBCO 2023 bathymetric data, sea-level eustasy curves from published literature, and palaeoenvironmental models for vegetation and hydrology.
Data Pipeline
The reconstruction pipeline starts with raw GEBCO 15-arc-second bathymetric tiles, reprojected to ETRS89 and cropped to the North Sea basin. Sea-level offsets are applied per time slice, and the resulting DEMs are imported into Blender via a custom Python bridge that preserves metric scale and georeferences the terrain.
River networks are derived hydraulically using GRASS GIS's r.watershed module, then validated against published archaeological and geological studies of Holocene drainage patterns.
Rendering & Visualisation
Each time slice is rendered in Blender with physically based materials: procedural terrain shading driven by slope, altitude and proximity to water; volumetric atmosphere adapted to the palaeoclimate of each period; and particle-based vegetation distribution matching habitat zone maps.
The final output is a series of still renders and a short animated sequence showing the progressive inundation from 10,000 BP to 6,000 BP.
