Integrating physical principles based modeling, efficient numerical and computational and data analytics methodology can lead to the long sought goal of truly predictive modeling with quantified errors and uncertainties and simulation to support decision-making and dramatic new insights for critical applications ranging from volcanic activity to traumatic brain injury. These results build on decades of work on a number of methodologies – high quality physical models, adaptive mesh refinement and other appropriate discretization methods, parallel computing, statistical surrogates, uncertainty quantification and propagation and a host of computing scheme improvements. This talk will present a series of developments in our group that have now led to what we believe are truly useable simulations. These then form the basis for developing cyberinfrastructure that advances scientific discovery and innovation for entire communities.

The research described enables new exciting curricula and I will also briefly describe efforts at putting together new doctoral and Masters programs to train students in these areas.

Bio:
Abani Patra is a Computational and Data Scientist and directs the newly constituted Institute for Computing and Data Sciences and University at Buffalo, SUNY at Buffalo, NY. He has a PhD in Computational and Applied Mathematics from the University of Texas-Austin. He has published over 100 peer reviewed articles and conference papers and is a Fellow of the US Association for Computational Mechanics. He has also directed programs at both the National Science Foundation and Department of Energy Office of Advanced Scientific Computing Research.