This content will become publicly available on Wed Jun 26 00:00:00 EDT 2024
Hierarchical off-diagonal low-rank approximation of Hessians in inverse problems, with application to ice sheet model initialization
Abstract
Obtaining lightweight and accurate approximations of discretized objective functional Hessians in inverse problems governed by partial differential equations (PDEs) is essential to make both deterministic and Bayesian statistical large-scale inverse problems computationally tractable. The cubic computational complexity of dense linear algebraic tasks, such as Cholesky factorization, that provide a means to sample Gaussian distributions and determine solutions of Newton linear systems is a computational bottleneck at large-scale. These tasks can be reduced to log-linear complexity by utilizing hierarchical off-diagonal low-rank (HODLR) matrix approximations. In this work, we show that a class of Hessians that arise from inverse problems governed by PDEs are well approximated by the HODLR matrix format. In particular, we study inverse problems governed by PDEs that model the instantaneous viscous flow of ice sheets. In these problems, we seek a spatially distributed basal sliding parameter field such that the flow predicted by the ice sheet model is consistent with ice sheet surface velocity observations. Here, we demonstrate the use of HODLR Hessian approximation to efficiently sample the Laplace approximation of the posterior distribution with covariance further approximated by HODLR matrix compression. Computational studies are performed which illustrate ice sheet problem regimes for which the Gauss–Newton data-misfit Hessianmore »
- Authors:
-
- Univ. of California, Merced, CA (United States)
- New York Univ. (NYU), NY (United States)
- Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
- Publication Date:
- Research Org.:
- Sandia National Lab. (SNL-NM), Albuquerque, NM (United States); Los Alamos National Laboratory (LANL), Los Alamos, NM (United States); Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States)
- Sponsoring Org.:
- USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR); USDOE Office of Science (SC), Biological and Environmental Research (BER); USDOE Office of Science (SC), Basic Energy Sciences (BES). Scientific User Facilities (SUF); USDOE National Nuclear Security Administration (NNSA); National Science Foundation (NSF)
- OSTI Identifier:
- 2311406
- Report Number(s):
- SAND-2023-08485J
Journal ID: ISSN 0266-5611
- Grant/Contract Number:
- NA0003525; AC02-05CH11231; DMS-1840265
- Resource Type:
- Accepted Manuscript
- Journal Name:
- Inverse Problems
- Additional Journal Information:
- Journal Volume: 39; Journal Issue: 8; Journal ID: ISSN 0266-5611
- Publisher:
- IOPscience
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 97 MATHEMATICS AND COMPUTING; Hessians; inverse problems; hierarchical matrices; HODLR matrices; ice-sheet models; large-scale
Citation Formats
Hartland, Tucker, Stadler, Georg, Perego, Mauro, Liegeois, Kim Anne J., and Petra, Noémi. Hierarchical off-diagonal low-rank approximation of Hessians in inverse problems, with application to ice sheet model initialization. United States: N. p., 2023.
Web. doi:10.1088/1361-6420/acd719.
Hartland, Tucker, Stadler, Georg, Perego, Mauro, Liegeois, Kim Anne J., & Petra, Noémi. Hierarchical off-diagonal low-rank approximation of Hessians in inverse problems, with application to ice sheet model initialization. United States. https://doi.org/10.1088/1361-6420/acd719
Hartland, Tucker, Stadler, Georg, Perego, Mauro, Liegeois, Kim Anne J., and Petra, Noémi. Mon .
"Hierarchical off-diagonal low-rank approximation of Hessians in inverse problems, with application to ice sheet model initialization". United States. https://doi.org/10.1088/1361-6420/acd719.
@article{osti_2311406,
title = {Hierarchical off-diagonal low-rank approximation of Hessians in inverse problems, with application to ice sheet model initialization},
author = {Hartland, Tucker and Stadler, Georg and Perego, Mauro and Liegeois, Kim Anne J. and Petra, Noémi},
abstractNote = {Obtaining lightweight and accurate approximations of discretized objective functional Hessians in inverse problems governed by partial differential equations (PDEs) is essential to make both deterministic and Bayesian statistical large-scale inverse problems computationally tractable. The cubic computational complexity of dense linear algebraic tasks, such as Cholesky factorization, that provide a means to sample Gaussian distributions and determine solutions of Newton linear systems is a computational bottleneck at large-scale. These tasks can be reduced to log-linear complexity by utilizing hierarchical off-diagonal low-rank (HODLR) matrix approximations. In this work, we show that a class of Hessians that arise from inverse problems governed by PDEs are well approximated by the HODLR matrix format. In particular, we study inverse problems governed by PDEs that model the instantaneous viscous flow of ice sheets. In these problems, we seek a spatially distributed basal sliding parameter field such that the flow predicted by the ice sheet model is consistent with ice sheet surface velocity observations. Here, we demonstrate the use of HODLR Hessian approximation to efficiently sample the Laplace approximation of the posterior distribution with covariance further approximated by HODLR matrix compression. Computational studies are performed which illustrate ice sheet problem regimes for which the Gauss–Newton data-misfit Hessian is more efficiently approximated by the HODLR matrix format than the low-rank (LR) format. We then demonstrate that HODLR approximations can be favorable, when compared to global LR approximations, for large-scale problems by studying the data-misfit Hessian associated with inverse problems governed by the first-order Stokes flow model on the Humboldt glacier and Greenland ice sheet.},
doi = {10.1088/1361-6420/acd719},
journal = {Inverse Problems},
number = 8,
volume = 39,
place = {United States},
year = {Mon Jun 26 00:00:00 EDT 2023},
month = {Mon Jun 26 00:00:00 EDT 2023}
}
Works referenced in this record:
A Fast Randomized Eigensolver with Structured LDL Factorization Update
journal, January 2014
- Xi, Yuanzhe; Xia, Jianlin; Chan, Raymond
- SIAM Journal on Matrix Analysis and Applications, Vol. 35, Issue 3
The creep of polycrystalline ice
journal, March 1955
- Glen, J. W.
- Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, Vol. 228, Issue 1175, p. 519-538
Finding Structure with Randomness: Probabilistic Algorithms for Constructing Approximate Matrix Decompositions
journal, January 2011
- Halko, N.; Martinsson, P. G.; Tropp, J. A.
- SIAM Review, Vol. 53, Issue 2
Consistent approximations and boundary conditions for ice-sheet dynamics from a principle of least action
journal, January 2010
- Dukowicz, John K.; Price, Stephen F.; Lipscomb, William H.
- Journal of Glaciology, Vol. 56, Issue 197
Data-sparse Approximation by Adaptive ℋ2-Matrices
journal, September 2002
- Hackbusch, W.; Börm, S.
- Computing, Vol. 69, Issue 1, p. 1-35
Mitigating the influence of the boundary on PDE-based covariance operators
journal, January 2018
- Daon, Yair; Stadler, Georg
- Inverse Problems & Imaging, Vol. 12, Issue 5
A Computational Framework for Infinite-Dimensional Bayesian Inverse Problems Part I: The Linearized Case, with Application to Global Seismic Inversion
journal, January 2013
- Bui-Thanh, Tan; Ghattas, Omar; Martin, James
- SIAM Journal on Scientific Computing, Vol. 35, Issue 6
Algorithms for Kullback--Leibler Approximation of Probability Measures in Infinite Dimensions
journal, January 2015
- Pinski, F. J.; Simpson, G.; Stuart, A. M.
- SIAM Journal on Scientific Computing, Vol. 37, Issue 6
An $$\mathcal O (N \log N)$$ O ( N log N ) Fast Direct Solver for Partial Hierarchically Semi-Separable Matrices: With Application to Radial Basis Function Interpolation
journal, April 2013
- Ambikasaran, Sivaram; Darve, Eric
- Journal of Scientific Computing, Vol. 57, Issue 3
A Computational Framework for Infinite-Dimensional Bayesian Inverse Problems, Part II: Stochastic Newton MCMC with Application to Ice Sheet Flow Inverse Problems
journal, January 2014
- Petra, Noemi; Martin, James; Stadler, Georg
- SIAM Journal on Scientific Computing, Vol. 36, Issue 4
Fast Algorithms for Bayesian Uncertainty Quantification in Large-Scale Linear Inverse Problems Based on Low-Rank Partial Hessian Approximations
journal, January 2011
- Flath, H. P.; Wilcox, L. C.; Akçelik, V.
- SIAM Journal on Scientific Computing, Vol. 33, Issue 1
A Fast Randomized Algorithm for Computing a Hierarchically Semiseparable Representation of a Matrix
journal, October 2011
- Martinsson, P. G.
- SIAM Journal on Matrix Analysis and Applications, Vol. 32, Issue 4
Spatial patterns of basal drag inferred using control methods from a full-Stokes and simpler models for Pine Island Glacier, West Antarctica: SPATIAL PATTERNS OF BASAL DRAG
journal, July 2010
- Morlighem, M.; Rignot, E.; Seroussi, H.
- Geophysical Research Letters, Vol. 37, Issue 14
PyAlbany: A Python interface to the C++ multiphysics solver Albany
journal, June 2023
- Liegeois, Kim; Perego, Mauro; Hartland, Tucker
- Journal of Computational and Applied Mathematics, Vol. 425
hIPPYlib-MUQ: A Bayesian Inference Software Framework for Integration of Data with Complex Predictive Models under Uncertainty
journal, June 2023
- Kim, Ki-Tae; Villa, Umberto; Parno, Matthew
- ACM Transactions on Mathematical Software, Vol. 49, Issue 2
A Sparse Matrix Arithmetic Based on $\Cal H$ -Matrices. Part I: Introduction to ${\Cal H}$ -Matrices
journal, April 1999
- Hackbusch, W.
- Computing, Vol. 62, Issue 2
Scalable and efficient algorithms for the propagation of uncertainty from data through inference to prediction for large-scale problems, with application to flow of the Antarctic ice sheet
journal, September 2015
- Isaac, Tobin; Petra, Noemi; Stadler, Georg
- Journal of Computational Physics, Vol. 296
Fast construction of hierarchical matrix representation from matrix–vector multiplication
journal, May 2011
- Lin, Lin; Lu, Jianfeng; Ying, Lexing
- Journal of Computational Physics, Vol. 230, Issue 10
Benchmark experiments for higher-order and full-Stokes ice sheet models (ISMIP–HOM)
journal, January 2008
- Pattyn, F.; Perichon, L.; Aschwanden, A.
- The Cryosphere, Vol. 2, Issue 2
Randomized GPU Algorithms for the Construction of Hierarchical Matrices from Matrix-Vector Operations
journal, January 2019
- Boukaram, Wajih; Turkiyyah, George; Keyes, David
- SIAM Journal on Scientific Computing, Vol. 41, Issue 4
MPAS-Albany Land Ice (MALI): a variable-resolution ice sheet model for Earth system modeling using Voronoi grids
journal, January 2018
- Hoffman, Matthew J.; Perego, Mauro; Price, Stephen F.
- Geoscientific Model Development, Vol. 11, Issue 9
Albany/FELIX : a parallel, scalable and robust, finite element, first-order Stokes approximation ice sheet solver built for advanced analysis
journal, January 2015
- Tezaur, I. K.; Perego, M.; Salinger, A. G.
- Geoscientific Model Development, Vol. 8, Issue 4
An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach: Link between Gaussian Fields and Gaussian Markov Random Fields
journal, August 2011
- Lindgren, Finn; Rue, Håvard; Lindström, Johan
- Journal of the Royal Statistical Society: Series B (Statistical Methodology), Vol. 73, Issue 4
Scalable Gaussian Process Computations Using Hierarchical Matrices
journal, September 2019
- Geoga, Christopher J.; Anitescu, Mihai; Stein, Michael L.
- Journal of Computational and Graphical Statistics, Vol. 29, Issue 2
On a Generalization of the Preconditioned Crank–Nicolson Metropolis Algorithm
journal, November 2016
- Rudolf, Daniel; Sprungk, Björn
- Foundations of Computational Mathematics, Vol. 18, Issue 2
Scalable Matrix-Free Adaptive Product-Convolution Approximation for Locally Translation-Invariant Operators
journal, January 2019
- Alger, Nick; Rao, Vishwas; Myers, Aaron
- SIAM Journal on Scientific Computing, Vol. 41, Issue 4
An inexact Gauss-Newton method for inversion of basal sliding and rheology parameters in a nonlinear Stokes ice sheet model
journal, January 2012
- Petrat, Noemi; Zhu, Hongyu; Stadler, Georg
- Journal of Glaciology, Vol. 58, Issue 211
Likelihood approximation with hierarchical matrices for large spatial datasets
journal, September 2019
- Litvinenko, Alexander; Sun, Ying; Genton, Marc G.
- Computational Statistics & Data Analysis, Vol. 137
Continental scale, high order, high spatial resolution, ice sheet modeling using the Ice Sheet System Model (ISSM): ICE SHEET SYSTEM MODEL
journal, March 2012
- Larour, E.; Seroussi, H.; Morlighem, M.
- Journal of Geophysical Research: Earth Surface, Vol. 117, Issue F1
A Bayesian approach to estimate uncertainty for full-waveform inversion using a priori information from depth migration
journal, September 2016
- Zhu, Hejun; Li, Siwei; Fomel, Sergey
- GEOPHYSICS, Vol. 81, Issue 5
Robust and Accurate Stopping Criteria for Adaptive Randomized Sampling in Matrix-Free Hierarchically Semiseparable Construction
journal, January 2019
- Gorman, Christopher; Chávez, Gustavo; Ghysels, Pieter
- SIAM Journal on Scientific Computing, Vol. 41, Issue 5
Fast computation of uncertainty quantification measures in the geostatistical approach to solve inverse problems
journal, August 2015
- Saibaba, Arvind K.; Kitanidis, Peter K.
- Advances in Water Resources, Vol. 82
Monte Carlo sampling methods using Markov chains and their applications
journal, April 1970
- Hastings, W. K.
- Biometrika, Vol. 57, Issue 1
Optimal initial conditions for coupling ice sheet models to Earth system models: PEREGO ET AL.
journal, September 2014
- Perego, Mauro; Price, Stephen; Stadler, Georg
- Journal of Geophysical Research: Earth Surface, Vol. 119, Issue 9
Compressing Rank-Structured Matrices via Randomized Sampling
journal, January 2016
- Martinsson, Per-Gunnar
- SIAM Journal on Scientific Computing, Vol. 38, Issue 4
Optimal Low-rank Approximations of Bayesian Linear Inverse Problems
journal, January 2015
- Spantini, Alessio; Solonen, Antti; Cui, Tiangang
- SIAM Journal on Scientific Computing, Vol. 37, Issue 6