Submerse: Visualizing Storm Surge Flooding Simulations in Immersive Display Ecologies

Area of Science:

• Computer Science
• Atmospheric Science
• Geographic Information Systems

Background:

• Effective visualization of complex flood simulation data is crucial for disaster management and urban planning.
• Existing visualization tools often struggle with large datasets and immersive environments, limiting their practical application.

Purpose of the Study:

• To develop and evaluate Submerse, an end-to-end framework for visualizing flood scenarios on large, immersive display systems.
• To enhance decision-making for emergency managers and stakeholders through interactive and detailed flood visualizations.

Main Methods:

• Reconstruction of surface meshes from flood simulation data and integration with geographical information (terrain, buildings).
• Optimization techniques including adaptive grids (dynamic quadtrees) and level-of-detail rendering for large datasets.
• Development of novel interaction methods: automatic viewpoint navigation and an AR-based focus+context technique.

Main Results:

• Successful reconstruction of to-scale 3D virtual scenes incorporating diverse geographical data.
• Demonstrated optimization for computation and memory performance with large simulation datasets.
• Validated effectiveness through workshops with emergency managers and domain experts using a superstorm flooding scenario.

Conclusions:

• Submerse provides an effective end-to-end solution for visualizing complex flood simulations in immersive environments.
• The framework's novel interaction techniques significantly improve the analysis and decision-making capabilities for disaster management.
• The system's utility was confirmed in realistic scenarios, highlighting its potential for real-world applications.