Abstract
We propose a spectral turning-bands approach for the simulation of second-order stationary vector Gaussian random fields. The approach improves existing spectral methods through coupling with importance sampling techniques. A notable insight is that one can simulate any vector random field whose direct and cross-covariance functions are continuous and absolutely integrable, provided that one knows the analytical expression of their spectral densities, without the need for these spectral densities to have a bounded support. The simulation algorithm is computationally faster than circulant-embedding techniques, lends itself to parallel computing and has a low memory storage requirement. Numerical examples with varied spatial correlation structures are presented to demonstrate the accuracy and versatility of the proposal.
Similar content being viewed by others
References
Alabert F (1987) The practice of fast conditional simulations through the LU decomposition of the covariance matrix. Math Geol 19(5):369–386
Almeida AS, Journel AG (1994) Joint simulation of multiple variables with a Markov-type coregionalization model. Math Geol 26(5):565–588
Arroyo D, Emery X, Peláez M (2012) An enhanced Gibbs sampler algorithm for non-conditional simulation of Gaussian random vectors. Comput Geosci 46:138–148
Bevilacqua M, Hering AS, Porcu E (2015) On the flexibility of multivariate covariance models. Stat Sci 30(2):167–169
Bochner S (1933) Monotone Funktionen Stieljessche integrale and harmonische analyse. Math Ann 108:378–410
Box GEP, Jenkins GM (1976) Time series analysis: forescasting and control, revised edn. Holden-Day, Oakland
Chan G, Wood A (1999) Simulation of stationary Gaussian vector fields. Stat Comput 9(4):265–268
Chilès JP, Delfiner P (1997) Discret exact simulation by the Fourier method. In: Baafi EY, Schofield NA (eds) Geostatistics Wollongong ’96, vol 1. Kluwer, Dordrecht, pp 258–269
Chilès JP, Delfiner P (2012) Geostatistics: Modeling Spatial Uncertainty, 2nd edn. Wiley, New York
Christakos G (1992) Random field models in earth sciences. Academic Press, San Diego
Cramér H (1940) On the theory of Stationary random functions. Ann Math 41:215–230
Daley DJ, Porcu E, Bevilacqua M (2015) Classes of compactly supported covariance functions for multivariate random fields. Stoch Environ Res Risk Assess 29(4):1249–1263
Davis MW (1987) Production of conditional simulations via the LU triangular decomposition of the covariance matrix. Math Geol 19(2):91–98
De Iaco S, Palma M, Posa D (2003) Covariance functions and models for complex-valued random fields. Stoch Environ Res Risk Assess 17(3):145–156
Dietrich CR, Newsam GN (1993) A fast and exact method for multidimensional Gaussian stochastic simulations. Water Resour Res 29(8):2861–2869
Emery X (2007) Reducing fluctuations in the sample variogram. Stoch Environ Res Risk Assess 21(4):391–403
Emery X (2008) A turning bands program for conditional Co-simulation of cross-correlated random fields. Comput Geosci 34(12):1850–1862
Emery X, Arroyo D, Peláez M (2014) Simulating large Gaussian random vectors subject to inequality constraints by Gibbs sampling. Math Geosci 46(3):265–283
Emery X, Lantuéjoul C (2006) TBSIM: a computer program for conditional simulation of three-dimensional Gaussian random fields via the turning bands method. Comput Geosci 32(10):1615–1628
Gaspari G, Cohn SE (1999) Construction of correlation functions in two and three dimensions. Q J R Meteorol Soc 125:723–757
Genton M, Kleiber W (2015) Cross-covariance functions for multivariate geostatistics. (With discussion). Stat Sci 30(2):147–163
Gneiting T, Kleiber W, Schlather M (2010) Matérn cross-covariance functions for multivariate random fields. J Am Stat Assoc 105(491):1167–1177
Gneiting T, Ševčíhová H, Percival DB, Schlather M, Jiang Y (2006) Fast and exact simulation of large Gaussian lattice systems in \(\mathbb{R}^2\): exploring the limits. J Comput Graph Stat 15(3):483–501
Golub G, Van Loan C (1989) Matrix computations. John Hopkins Press, Baltimore
Gradshtein IS, Ryzhik IM (1965) Table of integrals, series and products, 4th edn. Academic Press, New York
Grzebyk M, Wackernagel H (1994) Multivariate analysis and spatial/temporal scales: Real and complex Models. In Proceedings of XVIIth international biometric conference, Hamilton, Ontario vol 1 pp 19–33
Journel AG, Huijbregts CJ (1978) Mining geostatistics. Academic Press, London
Lantuéjoul C (2002) Geostatistical simulation: models and algorithms. Springer-Verlag, Berlin
Mantoglou A (1987) Digital simulation of multivariate two- and three-dimensional stochastic processes with a spectral turning bands method. Math Geol 19(2):129–149
Marcotte D (2012) Revisiting the linear model of coregionalization. In: Abrahamsen P, Hauge R, and Kolbjørnsen O, (eds). Proceedings of IX international geostatistics congress, Springer-Verlag, Oslo, 11–15 June 2012, pp 67–78
Matheron G (1973) The intrinsic random functions and their applications. Adv Appl Probab 5:439–468
Mignolet MP, Spanos PD (1992) Simulation of homogeneous two-dimensional random fields: part I-AR and ARMA models. J Appl Mech 59(2):260–269
Pardo-Iguzquiza E, Chica-Olmo M (1993) The fourier integral method: an efficient spectral method for simulation of random fields. Math Geol 25(2):177–217
Porcu E, Gregori P, Mateu J, Ostoja-Starzewski M (2012) New classes of spectral densities for lattice processes and random fields built from simple univariate margins. Stoch Environ Res Risk Assess 26(4):479–492
Ripley BD (1987) Stochastic Simulation. Wiley, New York
Schlather M (2012) Construction of covariance functions and unconditional simulation of random fields. Advances and Challenges in Space-Time modelling of natural events. Springer-Verlag, Lecture Notes in Statistics, pp 25–54
Shinozuka M (1971) Simulation of multivariate and multidimensional random processes. J Acoust Soc Am 49(1B):357–367
Shinozuka M, Jan CM (1972) Digital simulation of random processes and its applications. J Sound Vib 25(1):111–128
Spanos PD, Mignolet MP (1992) Simulation of homogeneous two-dimensional random fields: part II-MA and ARMA models. J Appl Mech 59(2):270–277
Ver Hoef JM, Barry RP (1998) Constructing and fitting models for cokriging and multivariate spatial prediction. J Stat Plan Infer 69:275–294
Wackernagel H (2003) Multivariate geostatistics: an introduction with applications. Springer-Verlag, Berlin
Wood ATA, Chan G (1994) Simulation of stationary Gaussian processes in \([0,1]^d\). J Comput Graph Stat 3(4):409–432
Zastavnyi VP (1991) Positive-definite functions depending on the norm. Solution of the Schoenberg problem. Institute of Applied Mathematics and Mechanics, Ukranian Academy of Sciences, Donetsk
Acknowledgments
The authors acknowledge the funding by the Chilean Commission for Scientific and Technological Research, through Projects CONICYT / FONDECYT / REGULAR / No. 1130085, CONICYT / FONDECYT / POSTDOCTORADO / No. 3140568 and CONICYT / FONDECYT / REGULAR / No. 1130647, respectively.
Author information
Authors and Affiliations
Corresponding author
Appendix: proof of equations (14), (16) and (17)
Appendix: proof of equations (14), (16) and (17)
Let \(Y_0\) be a scalar random field. The random field regularized by a sampling function \(\omega _r(\cdot )\) being the indicator function of the ball of \({\mathbb{R}}^d\) with arbitrary radius r is defined as (Chilès and Delfiner 2012)
If \(Y_0\) is a second-order stationary random field with covariance \(C_0({\mathbf{h}})\), then \(Y_{r}\) is also a second-order stationary random field, therefore one can define the following covariances that depend only on \({\mathbf{h}}\).
The covariance between \(Y_0({\mathbf{x}}+{\mathbf{h}})\) and \(Y_{r}({\mathbf{x}})\) is
that is
Similarly, the covariance between \(Y_r(\mathbf {x}+\mathbf {h})\) and \(Y_{r'}(\mathbf {x})\) is (Chilès and Delfiner 2012)
with \(\breve{\omega }({\mathbf{t}})=\omega (-{\mathbf{t}})\).
Since the Fourier transformation exchanges convolution and multiplication, Eqs. (18) and (19) in terms of Fourier transforms follow respectively, as:
where \(f_{00}\) is the spectral density of \(Y_0\), and \(\xi _{\omega _r}\), \(\xi _{\breve{\omega }_r}\) and \(\xi _{\omega _{r'}}\) are the Fourier transforms of \(\omega _r\), \(\breve{\omega }_r\) and \(\omega _{r'}\), respectively. As \(\omega _r\) is the indicator of a ball with radius r, one has \(\xi _{\omega _r} = \xi _{\breve{\omega }_r} = \xi (\cdot ,r)\) (Eq. (15)) (Gradshtein and Ryzhik 1965); likewise, \(\xi _{\omega _{r'}} = \xi (\cdot ,r')\).
Rights and permissions
About this article
Cite this article
Emery, X., Arroyo, D. & Porcu, E. An improved spectral turning-bands algorithm for simulating stationary vector Gaussian random fields. Stoch Environ Res Risk Assess 30, 1863–1873 (2016). https://doi.org/10.1007/s00477-015-1151-0
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00477-015-1151-0