Montana State researcher wins NSF grant to improve flood forecasting

BOZEMAN - A new grant from the National Science Foundation will help a Montana State University researcher improve the ability to predict floods resulting from rain falling on snow, like the 2022 flooding that pushed the Yellowstone River 14 feet over its banks, sweeping houses, highways and bridges into its roiling waters.

Siwei He, an assistant professor of civil engineering in the Norm Asbjornson College of Engineering at Montana State University, was recently awarded a $293,492 EPSCoR, or Established Program to Stimulate Competitive Research, grant from the NSF to enhance rain-on-snow weather forecasting.

Heavy rainfall alone doesn't account for the so-called 500-year flood - one of such magnitude that there's only a one-in-500 chance of it occurring in a given year, He said. That particular flooding event was worsened by rain that melted snowpacks in the Absaroka and Beartooth mountains, greatly amplifying the volume of water that rushed into the Yellowstone.

"Hydrologists currently struggle with making accurate, hyper-local predictions of rain-on-snow floods," He said. "The level of flooding of the Yellowstone River was a surprise to many."

At the time of that natural disaster, He was a researcher at the National Oceanic and Atmospheric Administration's Global Systems Laboratory at the University of Colorado-Boulder. A specialist in hydrological modeling - a way of simulating the movement, distribution and qualities of water using mathematics - He watched videos of the river raging beyond its banks to reshape parts of Paradise Valley and force a weeklong closure of Yellowstone National Park. Later, when He saw that MSU was accepting applications for a faculty member with hydrological modeling experience, he knew it would be a chance to research ways to predict more accurately where and when dangerous rain-on-snow floods might strike.

"Rain-on-snow floods are complex and difficult to predict, but they have serious implications for water management and disaster preparedness, particularly in mountainous and cold regions like Montana," He said. "We need to improve our capability to make accurate predictions to give people time to prepare for such disasters."

The NSF project involves collaborations with the Iowa Flood Center at the University of Iowa, which has a reputation for providing community-level flood forecasts to Iowans, and with NOAA National Weather Service Missouri Basin River Forecast Center, which provides operational streamflow forecasts for the Missouri Basin.

"Using this grant, I will develop numerical methods focused on simulating rain-on-snow streamflow," He said. "The model will be tested using historical rain-on-snow flooding events, such as the Yellowstone flood. If the performance is good, we will then be able to use the model to predict future rain-on-snow floods or streamflow."

The findings could inform future water resource management strategies and help mitigate risks associated with rain-on-snow floods in more communities. While the overall grant-funded study may be focused on a few specific rivers, He said the model could, one day, be adapted for implementation across all of Montana, providing hyper-resolution streamflow forecasting, such as predicting whether small creeks running through towns are expected to flood.

"These events are of increasing concern due to their potential impact on communities, water supplies and infrastructure, especially as climate change is projected to intensify such occurrences in many regions," He said.

"Better forecasts will help us to improve community preparedness."