An artificial intelligence system that spawns synthetic storms could help governments, businesses and the public prepare for risks under future climate scenarios.
Researchers and planners often use computerized, or synthetic, storms to help them predict risk, but these approaches have typically modeled storms using present-day conditions, not future climate scenarios. The AI technology developed by Ning Lin and graduate student Renzhi Jing generates synthetic hurricanes under both current conditions and those predicted as the planet warms.
Like previous storm generators, the new AI-based system, known as the Princeton environment-dependent probabilistic tropical cyclone (PepC) model, simulates the behavior of the storm as it germinates, moves over the ocean and makes landfall. But whereas most models in use today rely on relatively simple statistical tools and data from historical storms, the Princeton model utilizes machine learning, a subcategory of AI, in addition to storm and environmental data. Machine learning in this model involves algorithms that learn from studying the atmospheric and oceanic conditions from previous storms to predict the trajectory
and intensity of future storms.
The approach allows the storm’s intensification to evolve at different rates, from slow to moderate or rapid, as a response to environmental variables such as relative humidity, wind shear, and ocean conditions along the storm’s projected path. The synthetic-storm-generation model can be combined with climate models and other storm hazard models to project wind, storm surges and rainfall under various climate conditions. The results can help with engineering designs, updating building codes and zoning regulations, making flood maps, estimating insurance costs, conducting cost-benefit analyses and optimizing resources.
"If we can predict the potential consequences of climate change on storms and their associated hazards then we can make better decisions in terms of risk mitigation and adaptation.” – Ning LIn
Innovator: Ning Lin, Associate Professor of Civil and Environmental Engineering
Co-inventor: Graduate student Renzhi Jing
Development status: Patent protection is pending. This technology is available
Funding: National Science Foundation and National Oceanic and Atmospheric Administration
Contact: For licensing enquiries, contact Prabhpreet Gill, Technology Licensing Associate.