Technische Universität Berlin

“Hydrologists Are Not Prepared for Major Flash Floods”

Professor Eva Nora Paton discusses the link between droughts and extreme rainfall as well as their consequences

The news on 14 July 2021 that 150 liters of rain per square meter was expected in North Rhine-Westphalia and Rhineland-Palatinate caused little enough concern among those not used to hearing such reports. For hydrologist Professor Eva Nora Paton, however, it set alarm bells ringing. To provide an idea of just how much rain this is, the average rainfall for Berlin for the whole of July 2021 was around 75 liters per square meter. Professor Paton, who is managing director of the Institute of Ecology and head of the Chair of Ecohydrology and Landscape Evaluation at Technische Universität Berlin, visited the site of the floods in their immediate wake. She herself is originally from Hagen, one of the areas hit by the flooding. Here, she discusses her thoughts and observations.

Professor Paton, we hear so many terms used in connection with the events of 14 July 2021 - flood disaster, inundation, flash flood, heavy rain. Can you give us a clearer idea of what these terms mean?

There are different types of floods. Firstly, there is classic river flooding that involves a concentration of large amounts of runoff water in rivers. This is mostly the result of heavy, continuous rain falling over a large area for six, ten or twelve hours or even one to three days. This leads to an excess of water in medium-sized rivers such as the Ahr or the Moselle. At risk here are the flood plains immediately adjacent to the rivers, which is why these areas should not be built on. Generally, this type of flooding can be forecast, with advanced warning given to the relevant state or federal authorities including the various flood control centers. This is what should also have happened in the Ahr region.

Then there are flash floods, an extreme type of flooding. Flash floods quickly transform small streams into torrential rain rivers. This occurs suddenly and unexpectedly, mostly in very small catchments covering just a few square kilometers. A great deal of precipitation falls in very short periods of time, ranging from just minutes to a few hours. This rainfall cannot percolate quickly enough and flows off the surface with great force. Flash floods carry debris with them, ripping gullies in paths. Steep, narrow catchments and soil sealing intensify their effects. Flash floods themselves usually have very short durations of just a few hours. They occur at the headwaters of catchments, on innocuous-looking streams, the kind that may even dry up completely in periods of drought. The acute danger they pose can only be predicted with a very short period of warning. Flash floods can occur anywhere. Our current knowledge does not enable us to predict exactly where and when this kind of torrential rainfall will occur.

Dresden was affected in 2002. Why not this year? Has the flooding shifted to other areas?

The flooding of major rivers such as the Elbe, the Oder or the Rhine can only occur when heavy rain falls over large areas. The flooding in July of this year occurred at local to regional level and was not sufficient to cause extreme swelling of the Rhine. As the area of precipitation moved across the western German federal states, it was sufficient to cause flooding of medium-sized rivers such as the Moselle. However, as the rainfall throughout the entire Rhine catchment was not so heavy, it had little impact here.


How did the flooding disaster in North Rhine-Westphalia and Rhineland-Palatinate differ to previous floods?

What was new about the events of July 2021 was the combination of flooding and flash floods. It rained heavily and for a very short period in a large contiguous area containing a number of smaller rivers. Unlike in 2002 and 2006, so much rain fell within just a few hours that the earth was only able to absorb a little of this, causing the water to flow off even more quickly than we witnessed in earlier events. This took the people living in these areas completely unawares. They had simply not expected the water to cascade down the slopes with such force within a few hours or in some cases just some ten minutes. The task now facing hydrologists is to simulate this dynamic and understand why everything happened so quickly.

Do you have a theory?

One cause is undoubtedly the extreme precipitation that occurred within the space of just a few hours. Another factor is the changing land use in some areas. When examining the catchment in Hagen, I was struck by the level of spruce mortality and deforestation. This is the case not only in some parts of North Rhine-Westphalia, but throughout the entire Central German Uplands Region. Much of the forestland in the headwaters had been cleared as a result of the 2018 drought and bark beetle infestation.

In the catchment around Hagen, the area I examined in detail, timber harvesters with wide excavator tires are used to transport root debris and heavy trucks to clear the trees. In order for these vehicles to get to where they are needed, gravel areas were created on forest roads. These were then hardened by the heavy vehicles, so that the water just rushed down the slopes and right into the residential neighborhoods, generating perfect runoff channels along the former forest roads. It dismayed me to see how the forest is no longer able to perform its role within the ecosystem as interceptor, decelerator, and buffer, and actually helps the water to flow away immediately.

What can we do?

The spruce forests in the area were planted in the 1950s and take about one hundred years to fully mature. This means that the plan was to cut them down in stages over the next 20 to 30 years. However, the drought conditions of the last years has made them vulnerable and they have been infested by bark beetles. To prevent the bark beetles spreading, the spruces are being felled and cleared as quickly as possible. This is the main problem. In the area that I examined, land use has changed over the last two or three years as a result of spruce mortality. I assume the situation is similar in the affected catchments in the Eifel, as the entire central uplands belt is witnessing an extreme loss of spruce trees. The areas are simply too large to be able to afford to completely re-forest them. So this is being done in stages. It will be five to ten years before we have forests able to perform the function of a reservoir.

The warning system has also been criticized. Is this justified?

Many fatalities could have been prevented if people had been warned or had at least known how to react to the heavy rain. The reporting chains didn’t work. The reports coming in were not recognized as an alarm. I have now seen warnings from the German weather service as well as private weather services which had forecast rainfall of up to 150 millimeters on a single day. For a hydrologist this can only mean one thing: the town will be flooded. The people living there probably heard this and thought “Ok. Heavy rain.” In fact, schools, kindergartens, doctors, etc. should have been warned, so that they could pass on the warning and cancel appointments and events to help keep people away from dangerous areas. There were patients attending doctor’s practices who said, “When I was at the doctor’s, the rain wasn’t so heavy.” Just an hour later the parking lot next door was completely under water.

Consideration is being given to creating risk and vulnerability maps for all populated areas. These should show where people are in need of special assistance, for example hospitals, kindergartens or homes for the elderly or where dangerous toxic substances are stored. This would make it possible to create a detailed list of who needs to be warned.

Was what happened unprecedented?

We have only had high-resolution precipitation time series covering the whole of Germany for twenty years. So, from a hydrometeorological point of view, we are unable to say exactly what happened in these streams prior to that. What we can say based upon what we now know is that, since records began, we have never had such heavy rainfall in Germany falling within just six hours over a large region. In the recent past, there have been comparable flash floods, for example in 2016 in Braunsbach in Baden-Württemberg or Simbach am Inn in Lower Bavaria, where small streams also completely destroyed a settled area. However, these cases did not involve a large, contiguous area being affected all at once as is now the case in North Rhine-Westphalia and Rhineland-Palatinate.

What is the significance of this?

In general, we are seeing an increase in heavy rain. At the moment, we are unable to say if extreme rainfall events such as we witnessed in July of this year are on the rise. However, if they were to repeatedly occur (for example, several times in succession in a single year), hydrologists and water management experts in Germany would be as little prepared as epidemiologists were for the coronavirus pandemic.

How are the recent floods affecting the work of your academic chair?

We are going to focus on areas of spruce deforestation. I would also like to make management of extreme climate situations an integral part of our teaching program. This includes the process of planning for a crisis as well as analyses of extreme weather, operational forecasting and risk disclosures. As part of our program in Ecology and Environmental Planning, we need to prepare our students to work at the interfaces between different areas of expertise in order to be able to manage disasters over the next ten to twenty years. We need people who can combine expertise in hydrology and climatology to deal with disasters such as droughts, forest fires, and heat waves as well as heavy rain. In the future, we want our students to be able to work in different roles within the reporting chain, including administration, and process and understand complex data in the shortest of times so as to interpret warnings correctly and recommend the right measures.


Thank you for the interview.

Interviewer: Christina Camier.


Prof. Dr.

Eva Paton

Ecohydrology & Landscape Evaluation