DYNAMIC DATA-DRIVEN INVERSION FOR TERASCALE SIMULATIONS: REAL-TIME IDENTIFICATION OF AIRBORNE CONTAMINANTS

In contrast to traditional terascale simulations that have known, fixed data inputs, dynamic data-driven (DDD) applications are characterized by unknown data and informed by dynamic observations. DDD simulations give rise to inverse problems of determining unknown data from sparse observations. The main difficulty is that the optimality system is a boundary value problem in 4D space-time, even though the forward simulation is an initial value problem. We construct special-purpose parallel multigrid algorithms that exploit the spectral structure of the inverse operator. Experiments on problems of localizing airborne contaminant release from sparse observations in a regional atmospheric transport model demonstrate that 17-million-parameter inversion can be effected at a cost of just 18 forward simulations with high parallel efficiency. On 1024 Alphaserver EV68 processors, the turnaround time is just 29 minutes. Moreover, inverse problems with 135 million parameters - corresponding to 139 billion total space-time unknowns - are solved in less than 5 hours on the same number of processors. These results suggest that ultra-high resolution data-driven inversion can be carried out sufficiently rapidly for simulation-based "real-time" hazard assessment.

[1]  J. King,et al.  Multilevel algorithms for ill-posed problems , 1992 .

[2]  Kevin R. Long,et al.  A variational finite element method for source inversion for convective-diffusive transport , 2003 .

[3]  Numerische,et al.  A wavelet multilevel method for ill-posed problems stabilized by Tikhonov regularization , 1997 .

[4]  William L. Briggs,et al.  A multigrid tutorial , 1987 .

[5]  George Biros,et al.  Parallel Lagrange-Newton-Krylov-Schur Methods for PDE-Constrained Optimization. Part I: The Krylov-Schur Solver , 2005, SIAM J. Sci. Comput..

[6]  T. Hughes,et al.  Streamline upwind/Petrov-Galerkin formulations for convection dominated flows with particular emphasis on the incompressible Navier-Stokes equations , 1990 .

[7]  Dacian N. Daescu,et al.  An adjoint sensitivity method for the adaptive location of the observations in air quality modeling , 2003 .

[8]  Thomas Kaminski,et al.  A coarse grid three-dimensional global inverse model of the atmospheric transport 1. Adjoint model and Jacobian matrix , 1999 .

[9]  William L. Briggs,et al.  A multigrid tutorial, Second Edition , 2000 .

[10]  Martin Hanke,et al.  Two-level preconditioners for regularized inverse problems I: Theory , 1999, Numerische Mathematik.

[11]  Barbara Kaltenbacher.  V-cycle convergence of some multigrid methods for ill-posed problems , 2003, Math. Comput..

[12]  George Biros,et al.  Parallel Lagrange-Newton-Krylov-Schur Methods for PDE-Constrained Optimization. Part II: The Lagrange-Newton Solver and Its Application to Optimal Control of Steady Viscous Flows , 2005, SIAM J. Sci. Comput..

[13]  P. Hansen.  Rank-Deficient and Discrete Ill-Posed Problems: Numerical Aspects of Linear Inversion , 1987 .

[14]  Wolfgang Hackbusch,et al.  Multigrid Methods for FEM and BEM Applications , 2004 .

[15]  Marek Behr,et al.  The effect of stabilization in finite element methods for the optimal boundary control of the Oseen equations , 2004 .