Multivariate type G Matérn stochastic partial differential equation random fields

David Bolin, Jonas Wallin

Research output: Contribution to journalArticlepeer-review

16 Scopus citations


For many applications with multivariate data, random-field models capturing departures from Gaussianity within realizations are appropriate. For this reason, we formulate a new class of multivariate non-Gaussian models based on systems of stochastic partial differential equations with additive type G noise whose marginal covariance functions are of Matérn type. We consider four increasingly flexible constructions of the noise, where the first two are similar to existing copula-based models. In contrast with these, the last two constructions can model non-Gaussian spatial data without replicates. Computationally efficient methods for likelihood-based parameter estimation and probabilistic prediction are proposed, and the flexibility of the models suggested is illustrated by numerical examples and two statistical applications.
Original languageEnglish (US)
Pages (from-to)215-239
Number of pages25
JournalJournal of the Royal Statistical Society. Series B: Statistical Methodology
Issue number1
StatePublished - Dec 17 2019

Bibliographical note

KAUST Repository Item: Exported on 2020-10-01
Acknowledgements: This work has been supported by the Swedish Research Council under grant 2016-04187 and the Knut and Alice Wallenberg Foundation (grant KAW 20012.0067). The authors thank Holger Rootzén, the Joint Editor, Associate Editor and the reviewers for valuable comments on the manuscript. We also thank Mikael Kuusela for helping with the Argo data.


Dive into the research topics of 'Multivariate type G Matérn stochastic partial differential equation random fields'. Together they form a unique fingerprint.

Cite this