Document Type
Article
Publication Date
9-2021
Publisher
American Institute of Mathematical Sciences
Source Publication
Foundations of Data Science
Source ISSN
2639-8001
Abstract
Many recent advances in sequential assimilation of data into nonlinear high-dimensional models are modifications to particle filters which employ efficient searches of a high-dimensional state space. In this work, we present a complementary strategy that combines statistical emulators and particle filters. The emulators are used to learn and offer a computationally cheap approximation to the forward dynamic mapping. This emulator-particle filter (Emu-PF) approach requires a modest number of forward-model runs, but yields well-resolved posterior distributions even in non-Gaussian cases. We explore several modifications to the Emu-PF that utilize mechanisms for dimension reduction to efficiently fit the statistical emulator, and present a series of simulation experiments on an atypical Lorenz-96 system to demonstrate their performance. We conclude with a discussion on how the Emu-PF can be paired with modern particle filtering algorithms.
Recommended Citation
Maclean, John and Spiller, Elaine T., "A Surrogate-Based Approach to Nonlinear, non-Gaussian Joint State-Parameter Data Assimilation" (2021). Mathematical and Statistical Science Faculty Research and Publications. 102.
https://epublications.marquette.edu/math_fac/102
Comments
Accepted version. Foundations of Data Science, Vol. 3, No. 3 (September 2021): 589-614. DOI. © 2021 American Institute of Mathematical Sciences. Used with permission.