Climate change has led to an increase in extreme weather events over the past few decades. As a result, there’s a growing need for visualization systems that can help scientists, emergency managers, and other concerned stakeholders to better prepare for potential catastrophic events. Since, of all these disasters, floods occur the most frequently (they account for 43% of natural disasters worldwide), predicting how they may impact a certain region has become vital to human survival.
Aware of the challenge, Stony Brook University researchers Saeed Boorboor, Yoonsang Kim, Prof. Brian A. Colle, and Distinguished Professor and former Chair of Computer Science Arie E. Kaufman, along with BU researchers Ping Hu and Josef M. Moses, have created Submerse—a novel 3D immersive visualization system that can help stakeholders get a better understanding of the effects of an oncoming flood, and make decisions to better minimize the impact of a disaster.
Submerse solves multiple problems faced by existing systems. To start with, it successfully couples all the existing geographical data—like buildings, bridges, and roads—with storm surge flooding simulations really well. As a result, all the simulated events are synchronized in real-time, and in a way that is modular and scalable.
Secondly, the system solves a major problem in existing simulations, which are 2D and available on desktops with finite screens, requiring users to zoom and pan to study their area of interest during a catastrophic event. Through Submerse, stakeholders can get a bird’s eye view of the effects of the flood on an area by looking at screens that are spread across an entire room. The immersive environment not only gives them a better understanding of what is happening but the hand-held controllers and gestures available along with it allow for more acute investigation. Moreover, multiple people can work collaboratively, focusing on specific parts of the city and coming together to draw more effective solutions for mitigation.
The system also allows for stakeholders to predefine which areas are of most interest to them, and hence focuses only on these spaces once they start running the simulation.
Once the system was ready, the team conducted a specialized workshop for domain experts, inviting 10 participants with backgrounds in research and academia, NYC emergency managers, and National Weather Services (NWS). They unanimously commented that the 3D visualization and sense of realism, coupled with immersion, positively transformed the way they planned for preparation, evacuation, and resilience measures, in contrast to currently adopted 2D methods.
One participant reflected, saying, “The display of flooding in 3D with actual buildings was a much more effective method of communicating risks, rather than a numerical forecast or a 2D risk map.” Another participant commented that the visualizations felt like “a helicopter ride during a catastrophe where [they] get a wholesome view of the city, and then go down at street level to investigate its intensity.”
The feedback given by these participants enabled the team to not only understand the significance of their contributions but also plan what features they want to add to Submerse, including integrating mitigation tools that will allow stakeholders to view the outcomes of their decisions in real-time, developing a VR headset so viewers can see the simulation no matter where they are (instead of going to a specific room in a specific building), and adding functionality for remote and on-site stakeholders to collaborate so they can mitigate the effects of a flood more effectively.
Saeed Boorboor, Principal Research Scientist at SBU, shared his thoughts on the impact of the project, “Submersive is an end-to-end system that packages everything we might need to visualize storm surge flooding simulations, and presents it in an easy-to-interact-with 3D format. We’re looking forward to continue working on it and using it to mitigate a superstorm flood scenario in New York City and beyond.”
Ankita Nagpal
Communications Assistant