In a warming world, accurate and timely data on flooding can make a critical difference in disaster preparedness. Informatics doctoral student, Bekir Demiray, uses Artificial Intelligence (AI) to enhance existing flood datasets to limit the devastating effects of flooding. 

“Flooding is the number one natural disaster affecting humans in the world,” Demiray describes. “I hope that my research can improve flood forecasting to decrease casualties and lessen the economic consequences of flooding.” 

One way to enhance these datasets is to improve the temporal resolutions. For example, if the sensor network records the rainfall data every 10 minutes, Demiray works with the neural network to increase the data record to every five minutes instead. The augmented data increases the ability to predict flood disasters and allows for better disaster management.  

Dr. Ibrahim Demir is the Director of the University of Iowa Hydroinformatics Lab, which uses next generation information technologies, including AI and machine learning, for environmental engineering to enhance disaster preparedness. 

“As we see more and more flooding happening around the world, these data are critical, as is having accurate and timely forecasting,” Demir says. “Bekir’s models will be important, benefiting from machine learning and deep learning, providing us opportunities to achieve that goal.” 

Increasing awareness about flooding 

One important aspect of Demiray’s work is engineering accessible and interactive programs for the public to better understand the impact of flooding. To do this, Demiray and collaborators created a web-based, 3D game that helps people understand the risks of flooding and gives them the chance to take action to prevent it.  

“In the game, you can see how flooding would affect your area and then you can do mitigation actions to help prevent casualties and lessen the economic cost,” Demiray explains. “It is important to improve awareness of flooding for the public.”  

While this game is still in the beta-testing stage, Demiray says that it should be available for public use soon. In addition to interactive technologies to increase awareness, Demiray works with the Iowa Flood Center to connect with Iowa City’s student youth population and inform them about the causes and effects of flooding. 

“About three times a year, students from elementary and high schools come into our lab to learn about our work,” Demiray says. “It’s exciting because I can see learning happening and how the research we are doing influences these young students.” 

Impact for communities 

As a member of the Hydroinformatics Lab, Demiray is part of a large team of researchers working together across a variety of disciplines. Demiray describes that because of the sheer size and expertise of the group, there is ample support and opportunity for help from fellow lab members.  

“If there's ever something you don’t know, you can just ask the person next to you for help,” Demiray details. “In that way, I feel really lucky to be in this group.” 

Demir comments on Demiray’s role as a mentor and collaborator in this large lab operation. "Bekir is a really good team player,” Demir remarks. “He aids in onboarding new members by helping them to learn novel skills and figure out how to access certain resources. In addition, he's mentoring and supervising some younger students in our group.” 

Working as a team and helping new members will be vital skills for Demiray as his research has the potential to involve collaborating and connecting with policymakers, educators, and community leaders. After graduating, Demiray hopes to stay in academia.  

“I find the environment of academia to be more relaxing and everyone is very friendly,” Demiray says. “Also, the feeling of submitting a new paper and having it be accepted is really rewarding. It gives you confidence and joy.” 

Demiray’s research and the work of the Hydroinformatics Lab has significant impact beyond academia alone. “Environmental monitoring is the first step for understanding the risk of the communities, and this is especially important for resource-restricted communities,” Demir narrates.  

As flooding continues at increasing rates, using AI can help communities predict, prepare, and mitigate devastation caused by flooding.