New way to predict river flood risk could help prepare for disasters

We may now be able to better predict which rivers are at risk of extreme flooding, even if they haven’t caused severe floods before.

Extreme flooding, such as the devastating 2021 floods in Germany that killed at least 196 people, is difficult to predict, says Stefano Basso at the Norwegian Institute for Water Research in Oslo. Typically, researchers only know a river is prone to extreme floods if this has happened in the past. “For this method to work, you have to wait for something bad to happen,” says Basso.

To look for other ways to predict extreme flood risk, Basso and his colleagues analysed discharge and flooding data from 101 rivers in Germany and the US going back at least 30 years.

They split the floods into two groups: those that were small in magnitude and expected for that river, and those that were extreme. The second group consisted of floods that were between 20 and 35 times bigger than a normal flood for a particular river.

The researchers studied the properties of the rivers that had experienced these extreme floods, including factors such as magnitude and frequency of rainfall and how quickly water appeared to flow from the surrounding regions to the river.

From this analysis, the researchers came up with two parameters that appeared to explain whether a river was likely to cause extreme flooding or not: how the area draining to the river (known as the river basin) retains and releases water, and the balance between how much rain the river receives and how much water evaporates from the region.

Basso and his colleagues then analysed data going back at least 10 years from 2519 rivers in Germany and the US, using these two parameters to determine whether a river was likely to cause extreme flooding or not.

They identified several rivers in Germany and the US that haven’t had extreme floods, but are thought to be at risk of them. In Germany, they include the Rems river in Baden-Württemberg and the Wörnitz and Vils rivers in Bavaria. In the US, rivers at risk of extreme flooding include the Baron Fork of the Illinois river and the Cowpasture river in Virginia. The team is currently using its model to predict which rivers are at risk of extreme flooding in the UK.

Being able to predict extreme floods is important, says Basso, because it is these floods in urban areas that officials are often unprepared for and cause the most damage.

“Hydrologists have known for some time that the most catastrophic floods are generated by different physical processes than smaller ‘normal’ floods, but the beauty of this study is that, for the first time, it identifies the places that are more likely to experience these disasters,” says Paul Bates at the University of Bristol, UK.