Predicting water availability amid climate change peaks perspective, research of ��’s Raczynski

STARKVILLE, Miss.—When Krzysztof Raczynski was working on his master’s degree in his native country of Poland, he was enthused to see a large waterfall during a trip to the Tatra mountains.

However, when he arrived, Raczynski discovered the waterfall had run dry amid an ongoing drought in the region.

“That’s when I started thinking in a broader perspective about droughts, and as I continued with my Ph.D., I started studying the trigger points for droughts and how they interact with the environment,” said Raczynski, now an assistant research professor at ��’s Geosystems Research Institute with the GEO Project.

Raczynski is developing new tools and strategies for predicting flood and drought levels, particularly in the Southeast U.S. He said there are some more or less known patterns of drought and flood event distribution in historic data, but how those patterns will continue amid a changing climate remains to be seen. Using streamflow data from the National Water Model and several other key data sets, he is developing the models to support better prediction of drought and flood in changing conditions.

“If we can predict those patterns, we can better mitigate and adapt to climate change and its hydrological impacts,” Raczynski said. “For example, when it rains a lot, there are techniques to collect that water, store it and then use it during the dry periods. The question is, how long will we need to store it? I am trying to figure out those flood and drought patterns in order to answer those questions and hopefully help communities better manage their water.”

Krzysztof Raczynski (Photo by Grace Cockrell)

Raczynski came to �� in 2021 as a postdoctoral researcher at the Northern Gulf Institute, where he spent most of his time developing the Harmonic Oscillator Seasonal-Trend, or HOST modeling software, which automatically analyzes and detects patterns in time-series data. That system has helped him analyze approximately 50 years of streamflow data to gain new insights.

Now as a GRI faculty member at the university, Raczynski plans to expand his focus as the models are refined. To Raczynski’s surprise, the Southeast is following similar patterns to his native Poland.

“When I was working on my Ph.D. in Poland, we had about 10 year cycles where there would be droughts for 10 years and then the water levels would be higher for the next 10 years,” Raczynski said. “It turns out the Southeastern U.S. goes through very similar cycles, although they are about seven years long. The more I dug into the cause of the pattern, I found there are a lot of processes accumulating in those seven years, so I am trying to untangle all of that. It was an amazing thing for me when I realized that I had already seen similar behavior in a completely different region of the world.”

Raczynski said that one challenge he is trying to overcome is that most of the current hydrologic models are addressing instantaneous streamflow conditions, but those only tell a portion of the hydrologic story. For example, for some regions even if rainfall decreases during a drought because of flowing groundwater from aquifers, the drought may not yet be affecting water movement. In the reverse situation, in some cases, excess rain may take more time to show up in streams if water levels in aquifers are low.

“Once we have those models that can simulate the long-term and short-term, let’s combine them with the instantaneous models, like the National Water Model or NextGEN, and build something that’s predicting the entire cycle, that includes every single component that is there, not only the streamflow reactions to current conditions,” he said.

For more on ��’s Geosystems Research Institute, visit .

�� is taking care of what matters. Learn more at .

��