A non-Gaussian ensemble filter for assimilating infrequent noisy observations

John Harlim, Brian R. Hunt

Research output: Contribution to journalArticlepeer-review

34 Scopus citations


We present a modified ensemble Kalman filter that allows a non-Gaussian background error distribution. Using a distribution that decays more slowly than a Gaussian allows the filter to make a larger correction to the background state in cases where it deviates significantly from the truth. For high-dimensional systems, this approach can be used locally. We compare this non-Gaussian filter to its Gaussian counterpart (with multiplicative variance inflation) with the three-dimensional Lorenz-63 model, the 40-dimensional Lorenz-96 model, and Molteni's SPEEDY model, a global model with ∼105 state variables. When observations are sufficiently infrequent and noisy, the non-Gaussian filter yields a significant improvement in analysis and forecast errors.

Original languageEnglish (US)
Pages (from-to)225-237
Number of pages13
JournalTellus, Series A: Dynamic Meteorology and Oceanography
Issue number2
StatePublished - Mar 2007

All Science Journal Classification (ASJC) codes

  • Oceanography
  • Atmospheric Science


Dive into the research topics of 'A non-Gaussian ensemble filter for assimilating infrequent noisy observations'. Together they form a unique fingerprint.

Cite this