This thesis presents GeoTwin, an open-source browser-native 3D digital twin platform for flood risk visualization and analysis in Small Island Developing States (SIDS), with Dominica as a use case. GeoTwin combines a rapid flood simulation service (FastFlood) with a WebGL-based 3D rendering pipeline to show flood scenarios and impacts to stakeholders.
From the client, users set up a simulation and request results from the FastFlood service. The platform downloads the resulting water-height GeoTIFF and processes it into a grid of flood polygons for near real-time visualization. Flood layers are rendered using the engine’s terrain-hugging pipeline and a dedicated water shader to show semi-transparent, terrain-conforming inundation surfaces over 3D context layers. An embedded impact analysis module sums up precomputed affected buildings per flood grid cell into impact indicators (e.g. buildings at risk, estimated occupants, economic value) with category reporting and simple threshold-based risk levels.
GeoTwin’s architecture prioritizes accessibility (no plugins, commodity hardware), transparency (open source implementation) and stakeholder-oriented design. The prototype shows an end-to-end workflow simulation request, raster processing, terrain-conforming rendering and impact analysis all in a web application. The contribution is a practical, lightweight framework to operationalize flood scenarios into a 3D environment, where 2D maps are insufficient for effective communication and planning.