Contact information:
Name: Kenneth Duru
Phone:    +49 152 1824 9716
Email:    kenneth.c.duru@gmail.com
 Address:  Munich University  
Dept. of Earth and Env. Sciences 
Theresientr. 41, 80333 Munich

Short bio: (see CV)
I am a computational and  applied mathematician. My undergraduate education was in Mathematics and Statistics at the University of Calabar, Nigeria. In 2005, I moved to Sweden for graduate studies, obtaining a M.Sc. degree in  Scientific Computing (major Mathematics) from The Royal Institute of Technology (KTH), Stockholm. In June 2012, I earned a PhD degree in Scientific Computing (major Numerical Analysis)  from Uppsala University, Sweden, under the supervision of Prof. Gunilla Kreiss. The title of my PhD thesis is: Perfectly matched layers and high order difference methods for wave equations. Since my graduation, I have spent more than three years at the Stanford  University  working with Prof. Eric Dunham. At Stanford University, my  research focused on developing simulation codes (WaveQLab3D), and high order accurate and provably stable finite difference methods for waves and earthquake rupture dynamics in 3D heterogeneous and geometrically complex Earth models. My recent research effort centers on the development of efficient and robust numerical methods for the  simulations of wave phenomena  on exascale supercomputers. I am currently a member of the EU funded consortium, ExaHyPE, dedicated to the development of an exascale hyperpolic PDE egine. 

Research Interests:
  • Collaborative research in computational geosciences: earthquake rupture dynamics, volcanology, full seismic waveform modeling and inversion, induced seismicity, hydraulic fracturing, tsunami modeling, etc.   
  • Partial differential equations
  • Initial boundary value problems
  • High order accurate and time-stable methods: finite difference methods, discontinuous Galerkin methods, spectral element methods.
  • Linear and nonlinear boundary and interface phenomena
  • Perfectly matched layers and absorbing layers
  • Provably accurate software for computational mechanics
  • Time stable hybrid solvers for multi-physics and multi-scale problems.
  • Inverse problems, uncertainty quantification and data assimilation.