Climate change often comes down to how it affects water, whether it’s for drinking, electricity generation, or how flooding affects people and infrastructure. To better understand these impacts, Oak Ridge National Laboratory water resources engineer Sudershan Gangrade is integrating knowledge ranging from large-scale climate projections to local meteorology and hydrology and using high-performance computing to create a holistic view of the future.
In one of his recent projects, Gangrade conducted national-scale hydrologic modeling to study climate change effects on federal hydropower resources. The project analyzed 132 federal facilities that make up about 40% of the nation’s hydropower generation capacity.
To better predict how those facilities will be affected by climate extremes, Gangrade and his fellow researchers in ORNL’s Environmental Sciences Division translated global climate change predictions to impacts on local hydrology. The assessment used several models to look at future streamflow, hydropower operations and reservoir evaporation. The multiple-model approach provides insights to help federal hydropower facilities prepare for uncertainties in a changing climate, and will soon be expanded to nonfederal stakeholders.
“The hydrologic cycle is intensifying, and we can expect more extremes in the future,” Gangrade said. “This study provides helpful data for hydropower operators so they can better manage their reservoirs.”
Solutions for an advanced flood warning system
In another project, Gangrade is helping create a global flood forecasting capability using a model called TRITON, or Two-dimensional Runoff and Inundation Toolkit for Operational Needs, for the U.S. Air Force. The tool, created by ORNL and Tennessee Technological University, incorporates precipitation and stream flow routing to simulate floods. The capability will be scaled to a 10-meter resolution, roughly the scale of a tennis court, to provide fine detail for researchers. Gangrade and colleagues have simulated several flood events such as Hurricane Harvey using TRITON, as well as other global flooding events.
Gangrade is also integrating machine learning methods to develop hybrid reservoir operation models. He is using machine learning to detect historical reservoir operation patterns and create better long-term water management simulations.
He will soon take on the role of principal investigator for a new project combining his expertise in hydroclimate assessments and inundation models to assess flood vulnerability for Department of Defense facilities. The ensemble-based flood modeling will be conducted using downscaled climate projections to help identify vulnerable infrastructures in current and future climate scenarios.
In the long term, Gangrade would like to see his efforts integrated into a real-time flood forecasting and early warning tool for the general population.
“Climate change is already leading to more extreme events. We’re likely to see an increase in flooding frequency as well as severity,” Gangrade said. “What we’re developing now at ORNL gives us a much better understanding of water dynamics in the future.”
The next step is to come up with a better warning system for society so people have more time to evacuate, for better emergency planning and response, and for solutions that can be put in place now to make communities more resilient, he said. “That’s where our inundation capability can come into play. It is a very efficient and accurate tool.”