Mathematical Geosciences

, Volume 47, Issue 7, pp 843–865

# A General Probabilistic Approach for Inference of Gaussian Model Parameters from Noisy Data of Point and Volume Support

• Thomas Mejer Hansen
• Knud Skou Cordua
• Klaus Mosegaard
Article

## Abstract

Methods that rely on Gaussian statistics require a choice of a mean and covariance to describe a Gaussian probability distribution. This is the case using for example kriging, sequential Gaussian simulation, least-squares collocation, and least-squares-based inversion, to name a few examples. Here, an approach is presented that provides a general description of a likelihood function that describes the probability that a set of, possibly noisy, data of both point and/or volume support is a realization from a Gaussian probability distribution with a specific set of Gaussian model parameters. Using this likelihood function, the problem of inferring the parameters of a Gaussian model is posed as a non-linear inverse problem using a general probabilistic formulation. The solution to the inverse problem is then the a posteriori probability distribution over the parameters describing a Gaussian model, from which a sample can be obtained using, e.g., the extended Metropolis algorithm. This approach allows detailed uncertainty and resolution analysis of the Gaussian model parameters. The method is tested on noisy data of both point and volume support, mimicking data from remote sensing and cross-hole tomography.

### Keywords

Covariance parameters Linear inversion Kriging  Collocation

### References

1. Arabelos D, Tscherning CC (2003) Globally covering a-priori regional gravity covariance models. Adv Geosci 1:143–147
2. Asli M, Marcotte D, Chouteau M (2000) Direct inversion of gravity data by cokriging. In: Kleingeld W, Krige D (eds) Proceedings of the 6th international geostatistics congress, Cape Town, South Africa, 10–14 April, pp 64–73Google Scholar
3. Atkinson PM (2013) Downscaling in remote sensing. Int J Appl Earth Obs Geoinf 22:106–114. doi:10.1016/j.jag.2012.04.012
5. Chiles J-P, Delfiner P (2012) Geostatistics: modeling spatial uncertainty. Wiley series in probability and statistics. Wiley, Hoboken, NJGoogle Scholar
6. Cordua KS, Looms MC, Nielsen L (2008) Accounting for correlated data errors during inversion of cross-borehole ground penetrating radar data. Vadose Zone J 7(1):263
7. Cressie N (1985) Fitting variogram models by weighted least squares. Math Geol 17(5):563–586
8. Desassis N, Renard P (2013) Automatic variogram modeling by iterative least squares: univariate and multivariate cases. Math Geosci 34(4):453–470
9. Emery X (2010) Iterative algorithms for fitting a linear model of coregionalization. Comput Geosci 36(9):1150–1160
10. Frykman P, Deutsch C (1999) Geostatistical scaling laws applied to core and log data. In: Proceedings of SPE annual technical conference and exhibition, pp 887–898Google Scholar
11. Frykman P, Deutsch C (2002) Practical application of geostatistical scaling laws for data integration. Petrophys 43(3):153–171Google Scholar
12. Georgii H-O (2008) Stochastics: introduction to probability and statistics, 1st edn. Walter de Gruyter, Berlin. ISBN: 3110191458Google Scholar
13. Giroux B, Gloaguen E, Chouteau M (2007) bh_tomo: a Matlab borehole georadar 2d tomography package. Comput Geosci 33(1):126–137
14. Gloaguen E, Marcotte D, Chouteau M, Perroud H (2005) Borehole radar velocity inversion using cokriging and cosimulation. J Appl Geophys 57(4):242–259
15. Gloaguen E, Marcotte D, Giroux B, Dubreuil-Boisclair C, Chouteau M, Aubertin M (2007) Stochastic borehole radar velocity and attenuation tomographies using cokriging and cosimulation. J Appl Geophys 62(2):141–157
16. Goovaerts P (1997) Geostatistics for natural resources evalutaion. Applied geostatistics series. Oxford University Press, New YorkGoogle Scholar
17. Goovaerts P (2008) Kriging and semivariogram deconvolution in the presence of irregular geographical units. Math Geosc 40(1):101–128
18. Goovaerts P (2010) Combining areal and point data in geostatistical interpolation: applications to soil science and medical geography. Math Geosc 42(5):535–554
19. Hansen TM, Mosegaard K (2008) VISIM: sequential simulation for linear inverse problems. Comput Geosci 34(1):53–76
20. Hansen TM, Journel AG, Tarantola A, Mosegaard K (2006) Linear inverse Gaussian theory and geostatistics. Geophysics 71(6):101–111
21. Hansen TM, Looms MC, Nielsen L (2008) Inferring the subsurface structural covariance model using cross-borehole ground penetrating radar tomography. Vadose Zone J 7(1):249–262
22. Hansen TM, Cordua KS, Looms MC, Mosegaard K (2013a) SIPPI: a Matlab toolbox for sampling the solution to inverse problems with complex prior information: part 1, methodology. Comput Geosci 52:470–480. doi:10.1016/j.cageo.2012.09.004
23. Hansen TM, Cordua KS, Looms MC, Mosegaard K (2013b) SIPPI: a Matlab toolbox for sampling the solution to inverse problems with complex prior information: part 2, application to cross hole GPR tomography. Comput Geosci 52:481–492. doi:10.1016/j.cageo.2012.10.001
24. Hansen TM, Cordua KS, Jacobsen BH, Mosegaard K (2014) Accounting for imperfect forward modeling in geophysical inverse problems exemplified for crosshole tomography. Geophysics 79(3):1–21
25. Hastings WK (1970) Monte Carlo sampling methods using Markov chains and their applications. Biometrika 57(1):97
26. Herzfeld UC (1992) Least-squares collocation, geophysical inverse theory and geostatistics: a bird’s eye view. Geophys J Int 111(2):237–249
27. Isaaks EH, Srivastava RM (1989) Applied geostatistics. Oxford University Press, OxfordGoogle Scholar
28. Jarmołowski W, Bakuła M (2014) Precise estimation of covariance parameters in least-squares collocation by restricted maximum likelihood. Studia Geophysica et Geodaetica 58(2):171–189
29. Jensen JM, Jacobsen BH, Christensen-Dalsgaard J (2000) Sensitivity kernels for time-distance inversion. Solar Phys 192(1–2):231–239
30. Journel AG, Huijbregts CJ (1978) Mining geostatistics. Academic Press, New YorkGoogle Scholar
31. Kay SM (2006) Intuitive probability and random processes using matlab. Springer, New York
32. Kelsall J, Wakefield J (2002) Modeling spatial variation in disease risk. J Am Stat Assoc 97(459):692–701
33. Kitanidis PK, Lane RW (1985) Maximum likelihood parameter estimation of hydrologic spatial processes by the Gauss–Newton method. J Hydrol 79(1/2):53–71
34. Knudsen P (1987) Estimation and modelling of the local empirical covariance function using gravity and satellite altimeter data. Bull Geod 61(2):145–160
35. Krarup T (1969) A contribution to the mathematical foundation of physical geodesy. Meddelse no. 44, Geodaetisk Institut, Koebenhavn, p 80Google Scholar
36. Kupfersberger H, Deutsch CV, Journel AG (1998) Deriving constraints on small-scale variograms due to variograms of large-scale data. Math Geol 30(7):837–852
37. Kyriakidis PC (2004) A geostatistical framework for area-to-point spatial interpolation. Geogr Anal 36(3):259–289
38. Lark R, Papritz A (2003) Fitting a linear model of coregionalization for soil properties using simulated annealing. Geoderma 115(3):245–260
39. Lele SR, Das A (2000) Elicited data and incorporation of expert opinion for statistical inference in spatial studies. Math Geol 32(4):465–487. doi:10.1023/A:1007525900030
40. Liu JS (1996) Metropolized independent sampling with comparisons to rejection sampling and importance sampling. Stat Comput 6(2):113–119
41. Liu Y, Journel AG (2009) A package for geostatistical integration of coarse and fine scale data. Comput Geosci 35(3):527–547
42. Looms MC, Hansen TM, Cordua KS, Nielsen L, Jensen KH, Binley A (2010) Geostatistical inference using crosshole ground-penetrating radar. Geophysics 75(6):29
43. Mosegaard K (1998) Resolution analysis of general inverse problems through inverse Monte Carlo sampling. Inverse Probl 14:405
44. Mosegaard K, Tarantola A (1995) Monte Carlo sampling of solutions to inverse problems. J Geophys Res 100(B7):12431–12447
45. Pardo-Igúzquiza E (1997) Mlreml: a computer program for the inference of spatial covariance parameters by maximum likelihood and restricted maximum likelihood. Comput Geosci 23(2):153–162
46. Pardo-Igúzquiza E (1998) Maximum likelihood estimation of spatial covariance parameters. Math Geol 30(1):95–108
47. Pardo-Igúzquiza E (1999) Varfit: a Fortran-77 program for fitting variogram models by weighted least squares. Comput Geosci 25(3):251–261
48. Pardo-Igúzquiza E, Dowd PA (2005) Emlk2d: a computer program for spatial estimation using empirical maximum likelihood kriging. Comput Geosci 31:361–370
49. Remy N, Boucher A, Wu J (2008) Applied geostatistics with SGeMS: a user’s guide. Cambridge University Press, CambridgeGoogle Scholar
50. Shamsipour P, Marcotte D, Chouteau M, Keating P (2010) 3d stochastic inversion of gravity data using cokriging and cosimulation. Geophysics 75(1):1–10
51. Tarantola A (2005) Inverse problem theory and methods for model parameter estimation. Society of Industrialand Applies Mathematics, PhiladelphiaGoogle Scholar
52. Tarantola A, Valette B (1982a) Generalized nonlinear inverse problems solved using the least squares criterion. Rev Geophys Space Phys 20(2):219–232
53. Tarantola A, Valette B (1982b) Inverse problems = quest for information. J Geophys 50(3):150–170Google Scholar
54. Truong PN, Heuvelink GMB, Pebesma E (2014) Bayesian area-to-point kriging using expert knowledge as informative priors. Int J Appl Earth Obs Geoinf 30:128–138. doi:10.1016/j.jag.2014.01.019
55. Warnes J, Ripley B (1987) Problems with likelihood estimation of covariance functions of spatial gaussian processes. Biometrika 74(3):640–642

© International Association for Mathematical Geosciences 2014

## Authors and Affiliations

• Thomas Mejer Hansen
• 1
• Knud Skou Cordua
• 1
• Klaus Mosegaard
• 1
1. 1.Niels Bohr InstituteUniversity of CopenhagenCopenhagenDenmark