Numerical strategies for filtering partially observed stiff stochastic differential equations

Research output: Contribution to journalArticlepeer-review

5 Scopus citations


In this paper, we present a fast numerical strategy for filtering stochastic differential equations with multiscale features. This method is designed such that it does not violate the practical linear observability condition and, more importantly, it does not require the computationally expensive cross correlation statistics between multiscale variables that are typically needed in standard filtering approach. The proposed filtering algorithm comprises of a " macro-filter" that borrows ideas from the Heterogeneous Multiscale Methods and a " micro-filter" that reinitializes the fast microscopic variables to statistically reflect the unbiased slow macroscopic estimate obtained from the macro-filter and macroscopic observations at asynchronous times. We will show that the proposed micro-filter is equivalent to solving an inverse problem for parameterizing differential equations. Numerically, we will show that this microscopic reinitialization is an important novel feature for accurate filtered solutions, especially when the microscopic dynamics is not mixing at all.

Original languageEnglish (US)
Pages (from-to)744-762
Number of pages19
JournalJournal of Computational Physics
Issue number3
StatePublished - Feb 1 2011

All Science Journal Classification (ASJC) codes

  • Numerical Analysis
  • Modeling and Simulation
  • Physics and Astronomy (miscellaneous)
  • Physics and Astronomy(all)
  • Computer Science Applications
  • Computational Mathematics
  • Applied Mathematics


Dive into the research topics of 'Numerical strategies for filtering partially observed stiff stochastic differential equations'. Together they form a unique fingerprint.

Cite this