Skip to main content
Loading Events

« All Events

  • This event has passed.

Graduate Defense: Brian Kyanjo

July 18 @ 3:00 pm - 5:00 pm

Dissertation Defense

Dissertation Information

Title: Geoflood: Computational Model For Overland Flooding

Program: Doctor of Philosophy in Computing

Advisor: Dr. Donna Calhoun, Mathematics

Committee Members: Dr. Jodi Mead, Mathematics; Dr. Michal Kopera, Mathematics; and Dr. David L. George, Computing

Abstract

Overland flooding, a critical environmental phenomenon, poses significant challenges for computational modeling due to its complex hydrodynamics and the need for high-resolution data. This thesis presents GeoFlood, a new open-source software package for overland flooding simulations. The computational model solves shallow water equations (SWE) on a quadtree hierarchy of mapped, logically Cartesian grids managed by the parallel, adaptive library ForestClaw (Calhoun & Burstedde, 2017a). The model is validated using standard benchmark tests from Neelz & Pender (2013) and against George (2011a) results obtained from the GeoClaw software (Clawpack Development Team, 2020) for the historical Malpasset dam break problem. The benchmark test results are compared against GeoClaw and a standard software HEC-RAS (Hy-draulic Engineering Center River Analysis System) results (Brunner, 2018). This comparison demonstrates GeoFlood’s ability to predict flood wave propagation accurately and efficiently on complex terrain. The results from comparisons with the Malpasset dam-break show good agreement with the GeoClaw results and are consistent with the historical records of the event. To enhance simulation speed, a hybrid CPU/GPU version of GeoFlood has been implemented, leveraging the parallelism provided by the ForestClaw library and GPU acceleration through CUDA programming. The primary goal of this research aspect is to accelerate existing CPU-based patch solvers and Riemann solvers within the GeoFlood code using CUDA. The accelerated version of GeoFlood has been validated against the CPU version, demonstrating substantial improvements in both advance time (time taken to advance a solution on a patch) and wall time. The findings indicate that the GPU-accelerated version of GeoFlood can simulate over- land flooding events more efficiently and accurately than its CPU counterpart. This GPU-accelerated version is anticipated to be a valuable resource for researchers and practitioners in hydrology and hydraulic engineering, enabling more precise and efficient simulations of overland flooding events.