A random map implementation of implicit filters

Abstract

Implicit particle filters for data assimilation generate high-probability samples by representing each particle location as a separate function of a common reference variable. This representation requires that a certain underdetermined equation be solved for each particle and at each time an observation becomes available. We present a new implementation of implicit filters in which we find the solution of the equation via a random map. As examples, we assimilate data for a stochastically driven Lorenz system with sparse observations and for a stochastic Kuramoto-Sivashinski equation with observations that are sparse in both space and time.

References in corpus (1)

• Sharp failure rates for the bootstrap particle filter in high dimensions

Cited by in corpus (24)

• An introduction to sampling via measure transport
• Three discussions of the paper "sequential quasi-Monte Carlo sampling", by M. Gerber and N. Chopin
• Transport map accelerated Markov chain Monte Carlo
• A mild Ito formula for SPDEs
• Scalable posterior approximations for large-scale Bayesian inverse problems via likelihood-informed parameter and state reduction
• Limitations of polynomial chaos expansions in the Bayesian solution of inverse problems
• Implicit particle methods and their connection with variational data assimilation
• Data Assimilation: The Schrödinger Perspective
• Implicit particle filtering for models with partial noise, and an application to geomagnetic data assimilation
• An exponential Wagner-Platen type scheme for SPDEs
• Conditions for successful data assimilation
• Model and Data Reduction for Data Assimilation: Particle Filters Employing Projected Forecasts and Data with Application to a Shallow Water Model
• Parameter estimation by implicit sampling
• A McKean optimal transportation perspective on Feynman-Kac formulae with application to data assimilation
• Improving particle filter performance by smoothing observations
• Ensemble filter techniques for intermittent data assimilation - a survey
• A Stochastic Maximum Principle Approach for Reinforcement Learning with Parameterized Environment
• Joint state-parameter estimation of a nonlinear stochastic energy balance model from sparse noisy data
• Scalable optimization-based sampling on function space
• Particle filters for high-dimensional geoscience applications: a review
• A Drift Homotopy Implicit Particle Filter Method for Nonlinear Filtering problems
• Symmetrized importance samplers for stochastic differential equations
• Optimization-Based MCMC Methods for Nonlinear Hierarchical Statistical Inverse Problems
• An acceleration strategy for randomize-then-optimize sampling via deep neural networks