Enhancement Digital Twin Application for Disaster Management by UNOOSA/UN-SPIDER

This presentation details the "Enhancement Digital Twin Application" initiative led by the United Nations Office for Outer Space Affairs (UNOOSA) and UN-SPIDER. Designed to support the United Nations' "Early Warnings for All" (EW4All) agenda, this project leverages cutting-edge space technologies to bolster disaster resilience in Small Island Developing States (SIDS). A primary challenge in the Pacific region is the significant "data gap" - specifically the lack of building footprints that include height information, which is critical for accurate disaster modeling. While global datasets exist, they often lack vertical data, and regional initiatives like PCRAFI have limited coverage. To bridge this gap, this project utilizes 30cm high-resolution satellite imagery combined with Deep Learning AI models to construct cost-effective 3D Digital Twins. The methodology employs advanced techniques, including NeRF and Gaussian Splatting, to generate models ranging from LOD1 (for GIS analysis) to LOD3 (for high-fidelity visualization). The core of the presentation focuses on the "Tonga Disaster Preparedness Platform," a pilot project implemented in 2024. This platform integrates these 3D geospatial models with real-time environmental data from IoT rain gauges and water-level sensors installed on the ground. This fusion enables precise, real-time simulations of sea-level rise and flood scenarios. A key innovation is the system's ability to optimize evacuation routes dynamically; by analyzing real-time flood depth data, the digital twin can identify safe passage corridors and update evacuation directions instantly, a capability that static hazard maps cannot provide. Finally, the presentation outlines the roadmap for expanding these capabilities to the Cook Islands and the Republic of Palau. It demonstrates how satellite-derived digital twins can revolutionize the entire Disaster Risk Management (DRM) cycle - spanning prevention, mitigation, response, and recovery - providing a scalable, data-driven framework for climate adaptation in vulnerable "big ocean" states.