ProyecTsu
Real-Time 3D Tsunami Visualization
From numerical tsunami models
to immersive real-time 3D environments.
Scientific Data Pipeline
The system is built around an automated data orchestration pipeline
designed to eliminate manual preprocessing and ensure reproducibility.
- GeoClaw – Physical tsunami simulation (fgout)
- Python Pipeline – Transcoding to RAW16 heightmaps
- Manifest Contract – Metadata-driven scene loading
- Unity Engine – GPU-based vertex displacement
Scientific Data Pipeline
The system is built around an automated data orchestration pipeline
designed to eliminate manual preprocessing and ensure reproducibility.
- GeoClaw – Physical tsunami simulation (fgout)
- Python Pipeline – Transcoding to RAW16 heightmaps
- Manifest Contract – Metadata-driven scene loading
- Unity Engine – GPU-based vertex displacement
Technical Validation
Validation was performed using the
eta_init_force_dry benchmark scenario from GeoClaw.
The reconstructed 3D visualization exhibits a
1:1 spatial correspondence
with the original numerical outputs.
Performance benchmarks demonstrate stable real-time rendering
above 30 FPS on consumer-grade GPUs,
validating the efficiency of the data transcoding strategy.
Visualization Results
Tsunami Surface Shader
The water surface deformation is computed using GPU-based vertex displacement:
η(x,y,t) = h(x,y,t) - h₀(x,y)
Where η represents the free surface elevation extracted from GeoClaw fgout data.
Technology Stack
Python
Unity 3D
GeoClaw
Rodrigo Cerda Calapuja
Desarrollador Principal – Ingeniería en Computación e Informática, Universidad Andrés Bello.
Diseño de arquitectura, integración GeoClaw–Python–Unity y desarrollo del MVP de escritorio.
Profesor Guía
Giannina Costa
Universidad Andrés Bello.
Supervisión académica, revisión del documento de tesis y alineamiento con los objetivos del proyecto de título.
Profesor Guía
Matías Vargas
Universidad Andrés Bello.
Supervisión académica, asesoría en guía técnica, arquitectura y funcionamiento de app.