FloodCitiSense: Predicting floods in an unpredictable world- with the help of citizen science…

Due to climate change, not only are extreme weather patterns becoming more common, they are also becoming more unpredictable. This combination of unpredictability and volatility poses significant risks to public health. One large area of concern is flooding. Extreme and unpredictable flooding could endanger millions of people living in cities with rivers. Last year alone, flash floods had severe impacts in France, Italy, Spain and the United Kingdom. Therefore, governments, water management authorities, and researchers must develop ever more sophisticated flood warning systems. FloodCitiSense tackles this problem head on, developing a sophisticated early warning system (EWS) for flooding to be used by inhabitants and municipal authorities across Europe. In this interview, project coordinator Boud Verbeiren shares results from Brussels, Rotterdam and Birmingham.


Innovation through cooperation

Boud starts by discussing what makes this project innovative compared to past projects. He identifies two elements. The first is the project’s emphasis on cocreation. They don’t think observing municipal actors and citizens is good enough. Instead, they are working with them to create their EWS. Cocreation makes sense because municipalities are already responsible for dealing with waterways and flood mitigation strategy, whilst citizens are the ones directly impacted by flooding. Boud says:

“Citizens should not be treated as passive victims of flooding,
instead they can play an important active role in mitigation.”

The second significant innovation was the implementation of the living lab approach to flood mitigation research. Boud explains this is an intensive methodology where researchers try to create experimental conditions in the real world. Implementing rigorous experimental standards in real world conditions can be extremely difficult because there many complexities to consider, but if done correctly the results are very rewarding. For example, testing a flooding EWS on a real flood inspires more confidence than and EWS tested on a simulation.


Credibility is key

One of the key concerns is ensuring the EWS generates accurate and reliable warnings. Boud tells us government actors are particularly concerned about this. He reasons, “they want to be really sure that the solution is reliable before promoting this to the wider public.” Boud concedes this will take more time in their case. At the end of the project, the system will only be visible to municipal actors and flood mitigation experts. Boud agrees that this is the right thing to do for now. Before this product can be given to the general public, it must go through many phases of rigorous testing and refinement to ensure it is credible enough for use by the general public. Boud believes they already have an excellent basis to develop their work into a fully operational EWS.

The fact that the project is being conducted on a European level also adds credibility. An international scope means more widely applicable results. Boud articulates how each city faces similar fluvial (river) flooding problems which are comparable, but they also have unique variables that are case specific. For example, Birmingham suffers from a higher risk of pluvial (rainfall) flooding. These differences are a key strength for two reasons: general patterns can be identified more confidently. On the other hand, the differences between the cities highlights the limitations of a general system and shows where context specific adjustments need to be made.


The power of citizen observatories  

Boud says the project has already fulfilled a key aim: making citizens active research partners. This was done through the creation of citizen observatories. Boud clarifies what that means: first, citizens were given cost efficient local sensors, which obtain important rainfall information in real time. Now, citizens are asked to input flooding impacts in the reporting app developed by the project. He says, “this kind of information is used in combination with the information from pre-existing sources, like a royal meteorological institute, to create a model for our EWS.”

One of the first problems the researchers faced was finding a way to motivate citizens to use the app regularly. After all, a flood isn’t exactly a common occurrence. Boud tells me “that’s why we came up with the idea of not just using the app for information about flooding but also rainfall.” This gives citizens an incentive to use the app more frequently as they can consult it at any time to see what rainfall levels in their neighbourhood. Boud adds how this is useful for the model they’re building: “If a heavy rainfall event occurs and there is no flooding, that’s also very important information for the modellers. We can then validate our model when it says there will be no flooding impact.”  By making it clear to citizens that input is still valuable in a non-flooding event, increases chances they will regularly offer it.


Predictions not prophecies will save us from the flood

Boud thinks they’ve created a great basis for eventually developing a push notification system (for mobiles) that warns people about imminent flooding. He admits that although they are still in a test phase, they have plans to scale up the project very soon. In their next test phase, beginning in April, they will make the app available to the wider public. Boud asserts, “this will give us even more useful data for the app and help us to make even finer predictions.” These predictions will eventually be good enough to send out potentially lifesaving push notifications.

Boud points to one last big idea that could save lives: creating a flood management dashboard for city managers. Currently all the project data is hosted in a central database, but Boud think it makes sense for city administrators to take control of the data for their city. It’s relatively easy to take the citizen observatory data and plug it into a dashboard. This would allow cities to create mutually beneficial feedback loops with their citizens. Citizens could help their city governments form a big picture view of flood risks. Then governments could more precisely warn their citizens of any imminent flood related danger. Boud finishes by saying:

 “We’ve created a way to get much more consistent information about pluvial flooding,
and we’ve raised public interest in flood research.”

> Contact FloodCitiSense




Please click here for the frequently asked questions we collected.
If you have an additional questions you are welcome to mail us at info@jpi-urbaneurope.eu