The SPDE approach for Gaussian and non-Gaussian fields: 10 years and still running

  title={The SPDE approach for Gaussian and non-Gaussian fields: 10 years and still running},
  author={Finn Lindgren and David Bolin and H. Rue},
  journal={Spatial Statistics},

Figures from this paper

A diffusion-based spatio-temporal extension of Gaussian Mat\'ern fields

This work develops a spatio-temporal extension of the Gaussian Mat ´ ern fields formulated as solutions to a stochastic partial differential equation that provides a sparse representation based on a sparse element approximation that is well suited for statistical inference and which is implemented in the R-INLA software.

The reproducing kernel Hilbert spaces underlying linear SDE Estimation, Kalman filtering and their relation to optimal control

It is often said that control and estimation problems are in duality. Recently, in Aubin-Frankowski (2021a), we found new reproducing kernels in Linear-Quadratic optimal control by focusing on the

Finite Element Representations of Gaussian Processes: Balancing Numerical and Statistical Accuracy

The theory implies that, under certain smoothness assumptions, one can reduce the computation and memory cost without hindering the estimation accuracy by setting 𝑛 ≪ 𝓁 in the large 𝐁 asymptotics.

The SPDE approach for spatio-temporal datasets with advection and diffusion

This work presents the advection-diffusion SPDE with first order derivative in time to enlarge the SPDE family to the space-time context and proposes Computationally efficient methods to estimate the parameters of the SPDe and to predict the spatio-temporalfield by kriging.

Realistic and Fast Modeling of Spatial Extremes over Large Geographical Domains

A more realistic Bayesian framework based on a novel Gaussian scale mixture model, where the Gaussian process component is defined by a stochastic partial differential equation that yields a sparse precision matrix, and the random scale component is modeled as a lowrank Pareto-tailed or Weibull-tailed spatial process determined by compactly supported basis functions.

Covariance-based rational approximations of fractional SPDEs for computationally efficient Bayesian inference

A new method based on approximating the covariance operator L − 2 β of the Gaussianfield u by a finite element method combined with a rational approximation of the fractional power results in a numerically stable GMRF approximation which can be combined with the integrated nested Laplace approximation (INLA) method for fast Bayesian inference.

Efficient algorithms for Bayesian Inverse Problems with Whittle-Matérn Priors

An efficient method for handling all admissible noninteger values of the exponent is derived and it is shown how to incorporate this prior representation into the infinite-dimensional Bayesian formulation, and how to compute the maximum a posteriori estimate, and approximate the posterior variance.

Numerical Approximation of Gaussian random fields on Closed Surfaces

The proposed numerical method relies on the Balakrishnan integral representation of the solution and does not require the approximation of eigenpairs and consists of a sinc quadrature coupled with a standard surface finite element method.

Boltzmann–Gibbs Random Fields with Mesh-free Precision Operators Based on Smoothed Particle Hydrodynamics

  • D. Hristopulos
  • Computer Science, Mathematics
    Theory of Probability and Mathematical Statistics
  • 2022
A new Boltzmann–Gibbs model which features local interactions in the energy functional and is embodied in a spatial coupling function which uses smoothed kernel-function approximations of spatial derivatives inspired from the theory of smoothed particle hydrodynamics.

Beyond Stationary Simulation; Modern Approaches to Stochastic Modelling

Three stochastic simulation methods using non-stationary covariance, multipoint simulation, and conditional GANs are presented: an SPDE method was used as a benchmark comparison and the application of machine learning techniques can be used to surpass previous simulation generation speeds and allows for greater parameterization.



A general framework for SPDE-based stationary random fields

This paper presents theoretical advances in the application of the Stochastic Partial Differential Equation (SPDE) approach in geostatistics. We show a general approach to construct stationary models

An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach

It is shown that, using an approximate stochastic weak solution to (linear) stochastically partial differential equations, some Gaussian fields in the Matérn class can provide an explicit link, for any triangulation of , between GFs and GMRFs, formulated as a basis function representation.

Spatial Matérn Fields Driven by Non‐Gaussian Noise

The article studies non‐Gaussian extensions of a recently discovered link between certain Gaussian random fields, expressed as solutions to stochastic partial differential equations (SPDEs), and

The SPDE Approach to Matérn Fields: Graph Representations

This paper investigates Gaussian Markov random field approximations to nonstationary Gaussian fields using graph representations of stochastic partial differential equations. We establish

Gaussian Markov Random Fields: Theory and Applications

This volume is essential reading for statisticians working in spatial theory and its applications, as well as quantitative researchers in a wide range of science fields where spatial data analysis is important.

Models and Methods for Random Fields in Spatial Statistics with Computational Efficiency from Markov Properties

  • D. Bolin
  • Computer Science, Mathematics
  • 2012
The SPDE method is extended to a larger class of non-Gaussian random fields with Matern covariance functions, including certain Laplace Moving Average (LMA) models, and it is shown how the SPDE formulation can be used to obtain an efficient simulation method and an accurate parameter estimation technique for a LMA model.

Multilevel approximation of Gaussian random fields: Covariance compression, estimation and spatial prediction

It is proved that the precision and covariance operators may be identified with bi-infinite matrices and finite sections may be diagonally preconditioned rendering the condition number independent of the dimension p of this section, and a tapering strategy by thresholding results in optimally numerically sparse approximations.

Matern Gaussian processes on Riemannian manifolds

This work proposes techniques for computing the kernels of these processes via spectral theory of the Laplace--Beltrami operator in a fully constructive manner, thereby allowing them to be trained via standard scalable techniques such as inducing point methods.