MATC hosted one webinar this semester thanks to Dr. Steven Corns, Associate Professor of Engineering Management and Systems Engineering at Missouri University of Science and Technology. His talk, “Deep Learning Techniques for Flash Flood Management” gave insight into his research finding a methodology that provides both risk quantification and optimal rerouting guidance for flash flooding.
The speed at which flash flooding occurs makes it both more dangerous and makes predicting its occurrence more challenging. Dr. Corns’ research uses publicly available geospatial data to create a historic flash flood database which can in turn be used to develop a deep learning model for classifying flash flood risk. Specifically, this project looks at elevation, slope, aspect, and curvature data for a subwatershed in Greene County, Missouri that frequently experiences flash floods.
Three machine learning models were analyzed and resulted in the artificial neural network exhibiting superior performance over logistic regression and support vector machine. Research also suggests future advancement in the capture of in-route traffic and road segments at high risk for flash flood event data for transportation officials to gain critical information to guide the deployment of resources more efficiently to reduce risk exposure to motorists.
To view more information on Dr. Corns, the webinar topic, and to view the recording, click here.
Share your News with MATC!
If you are a student, faculty member, or other affiliate of the Mid-America Transportation Center, we are eager to share your news of your work and accomplishments.
Send your information to Madison Schmidt at mschmidt24@unl.edu and it could appear in MATC's website, newsletter, Facebook, and Twitter.