What A River’s History Can Teach Us


Knowing the history of a river’s flow is surprisingly useful. It can help create policies for reservoir water release and help us understand events such as recurring droughts. But often existing records only go back so far.

Researchers at The City College of New York have developed a new model to recreate streamflow back before we started keeping track. The study, authored by Ph.D. student Arun Ravindranath and Professor Naresh Devineni, recreates the flow in the Upper Missouri River Basin back to the year 1800. The results appear in Water Resources Research.

“Let’s say a mega-drought event occurs every 80 or 100 years. If my streamflow record is only 62 years in length, how can I expect to understand this cycle?” said Ravindranath. “I might be tempted to conclude that this event is entirely a result of some other anomalous phenomena.”

While many reconstruction models recreate flow at individual points on a river, Ravindranath said, this new model considers the river as a network and looks at how water flows from one point to the next. Doing it this way increases the reconstruction’s accuracy and reduces uncertainty.

Their model gets help from nearby trees, too. By analyzing the width of tree rings, the scientists can get an idea of how much moisture was available that year. Wider rings indicate more moisture availability, while narrower rings mean less availability and thus a smaller volume of river flow.

“This model let us go, on average, more than 120 years back from when we have the first observations,” Ravindranath said.

Beyond SUM

Work By

Naresh Devineni ( Associate Professor , Civil Engineering) | Profile 1 | Profile 2
Arun Ravindranath (Ph.D. Candidate, Civil Engineering ) | Profile 1

Colleges and Schools

The City College of New York