Abstract
The Markov chain random field (MCRF) model/theory provides a non-linear spatial Bayesian updating solution at the neighborhood nearest data level for simulating categorical spatial variables. In the MCRF solution, the spatial dependencies among nearest data and the central random variable can be described by a probabilistic directed acyclic graph that conforms to a neighborhood-based Bayesian network on spatial data. By selecting different neighborhood sizes and structures, applying the spatial conditional independence assumption to nearest neighbors, or incorporating ancillary information, one may construct specific MCRF models based on the MCRF general solution for various application purposes. Simplified MCRF models based on assuming the spatial conditional independence of nearest data involve only spatial transition probabilities, and one can implement them easily in sequential simulations. In this article, we prove the spatial Bayesian network characteristic of MCRFs, and test the optimal neighborhoods under the spatial conditional independence assumption. The testing results indicate that the quadrantal (i.e., one nearest datum per quadrant) neighborhood is generally the best choice for the simplified MCRF solution, performing better than other sectored neighborhoods and non-sectored neighborhoods with regard to simulation accuracy and pattern rationality.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Banerjee, S., Carlin, B.P., Gelfand, A.E.: Hierarchical Modeling and Analysis for Spatial Data. CRC Press, Boca Raton (2004)
Bayes, T.: An essay towards solving a problem in the doctrine of chances. Phil. Trans. R. Soc. London. 53, 370–418 (1763)
Besag, J.: Spatial interaction and the statistical analysis of lattice systems (with discussions). J. R. Stat. Soc. Ser. B. 36(2), 192–236 (1974)
Besag, J.: On the statistical analysis of dirty pictures (with discussions). J. R. Stat. Soc. Ser. B. 48(3), 259–302 (1986)
Carle, S.F., Fogg, G.E.: Transition probability-based indicator geostatistics. Math. Geol. 28(4), 453–476 (1996)
Carle, S.F., Fogg, G.E.: Modeling spatial variability with one and multidimensional continuous-lag Markov chains. Math. Geol. 29(7), 891–918 (1997)
Charness, G., Levin, D.: When optimal choices feel wrong: a laboratory study of Bayesian updating, complexity, and affect. Am. Econ. Rev. 95(4), 1300–1309 (2005)
Christakos, G.: A Bayesian/maximum-entropy view to the spatial estimation problem. Math. Geol. 22(7), 763–777 (1990)
Dawid, A.P.: Conditional independence in statistical theory. J. R. Stat. Soc. Ser. B. 41(1), 1–31 (1979) J Roy Stat Soc Ser B
Dimitrakopoulos, R., Mustapha, H., Gloaguen, E.: High-order statistics of spatial random fields: exploring spatial cumulants for modeling complex non-Gaussian and non-linear phenomena. Math. Geosci. 42, 65–99 (2010)
Feller, W.: An Introduction to Probability Theory and its Applications, vol. 1, 2nd edn. John Wiley, New York (1957)
Friedman, N., Geiger, D., Goldszmidt, M.: Bayesian network classifiers. Mach. Learn. 29(2–3), 131–163 (1997)
Goovaerts, P.: Geostatistics for Natural Resources Evaluation. Oxford university press, New York (1997)
Haines-Young, R.: Exploring ecosystem service issues across diverse knowledge domains using Bayesian Belief Networks. Prog. Phys. Geogr. 35(5), 681–699 (2011)
Isaaks, E.H., Sarivastava, R.: An Introduction to Applied Geostatistics. Oxford University Press, New York (1989)
Jaffray, J.Y.: Bayesian updating and belief functions. IEEE Trans. Syst. Man Cyber. 22(5), 1144–1152 (1992)
Koltermann, E.C., Gorelick, S.M.: Heterogeneity in sedimentary deposits: a review of structure-imitating, process-imitating, and descriptive approaches. Water Resour. Res. 32, 2617–2658 (1996)
Li, W.: Markov chain random fields for estimation of categorical variables. Math. Geol. 39, 321–335 (2007a)
Li, W.: Transiograms for characterizing spatial variability of soil classes. Soil Sci. Soc. Am. J. 71, 881–893 (2007b)
Li, H., Reynolds, J.F.: A simulation experiment to quantify spatial heterogeneity in categorical maps. Ecology. 75(8), 2446–2455 (1994)
Li, W., Zhang, C.: A random-path Markov chain algorithm for simulating categorical soil variables from random point samples. Soil Sci. Soc. Am. J. 71, 656–668 (2007)
Li, W., Zhang, C.: A single-chain-based multidimensional Markov chain model for subsurface characterization. Environ. Ecol. Stat. 15(2), 157–174 (2008)
Li, W., Zhang, C.: Linear interpolation and joint model fitting of experimental transiograms for Markov chain simulation of categorical spatial variables. Int. J. Geogr. Info. Sci. 24(6), 821–839 (2010)
Li, Y., Yang, S., Liu, T., Dong, X.: Comparative assessment of semantic-sensitive satellite image retrieval: simple and context-sensitive Bayesian networks. Int. J. Geogr. Info. Sci. 26(2), 247–263 (2012)
Li, W., Zhang, C., Willig, M.R., Dey, D.K., Wang, G., You, L.: Bayesian Markov chain random field cosimulation for improving land cover classification accuracy. Math. Geosci. 47(2), 123–148 (2015)
Luo, J.: Transition probability approach to statistical analysis of spatial qualitative variables in geology. In: Forster, A., Merriam, D.F. (eds.) Proceedings of the 25th Anniversary Meeting of the International Association for Mathematical Geology (October 10–14, 1993, Prague, Czech Republic) and Geologic modeling and mapping, pp. 281–299. Plenum Press, New York (1996)
Luo, J., Thomsen, A.: On direct estimation of the bivariate probability distribution of a regionalized variable from its spatial samples. In: Fabbri, A.G., Royer, J.J. (eds.) The 3rd CODATA conference on geomathematics and geostatistics, vol. 32, pp. 115–123. Sci. de la Terre, Ser. Inf., Nancy (1994)
Matheron, G.: Principles of geostatistics. Econ. Geol. 58, 1246–1266 (1963)
Norberg, T., Rosen, L., Baran, A., Baran, S.: On modeling discrete geological structure as Markov random fields. Math. Geol. 34, 63–77 (2002)
Omre, H.: Bayesian kriging - merging observations and qualified guesses in kriging. Math. Geol. 19(1), 25–39 (1987)
Pearl, J.: Fusion, propagation, and structuring in belief networks. Artif. Intell. 29(3), 241–288 (1986)
Pearl, J.: Causality: Models, Reasoning, and Inference. Cambridge University Press, Cambridge (2000)
Pickard, D.K.: Unilateral Markov fields. Adv. Appl. Probab. 12(3), 655–671 (1980)
Ripley, B.D.: Gibbsian interaction models. In: Griffith, D.A. (ed.) Spatial Statistics: Past, Present, and Future, pp. 3–25. Institute of Mathematical Geography, Syracuse University, Syracuse (1990)
Rosholm, A. (1997) Statistical Methods for Segmentation and Classification of Images. Ph.D. dissertation, Technical University of Denmark, Lyngby
Schwarzacher, W.: The use of Markov chains in the study of sedimentary cycles. Math. Geol. 1, 17–39 (1969)
Strebelle, S.: Conditional simulation of complex geological structures using multiple-point statistics. Math. Geol. 34, 1–21 (2002)
Weissmann, G.S., Carle, S.F., Fogg, G.E.: Three-dimensional hydrofacies modeling based on soil surveys and transition probability geostatistics. Water Resour. Res. 35(6), 1761–1770 (1999)
Woodbury, A.D.: Bayesian updating revisited. Math. Geol. 21(3), 285–308 (1989)
Zhang, C., Li, W.: Regional-scale modeling of the spatial distribution of surface and subsurface textural classes in alluvial soils using Markov chain geostatistics. Soil Use Manag. 24(3), 263–272 (2008)
Zhang, W., Li, W., Zhang, C.: Land cover post-classifications by Markov chain geostatistical cosimulation based on pre-classifications by different conventional classifiers. Int. J. Remote Sens. 37(4), 926–949 (2016)
Zhang, W., Li, W., Zhang, C., Li, X.: Incorporating spectral similarity into Markov chain geostatistical cosimulation for reducing smoothing effect in land cover post-classification. IEEE J. Selected Topics Appl. Earth Obser. Remote Sens. 10(3), 1082–1095 (2017a)
Zhang, W., Li, W., Zhang, C., Ouimet, W.B.: Detecting horizontal and vertical urban growth from medium resolution imagery and its relationships with major socioeconomic factors. Int. J. Remote Sens. 38(12), 3704–3734 (2017b)
Acknowledgments
This manuscript was initially prepared in 2014. Partial support from U.S. NSF (Award numbers 1049017 and 1414108) is appreciated.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Disclaimer
Authors have the sole responsibility for all of the viewpoints presented in this paper.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
ESM 1
(DOCX 1114 kb)
Rights and permissions
About this article
Cite this article
Li, W., Zhang, C. Markov chain random fields in the perspective of spatial Bayesian networks and optimal neighborhoods for simulation of categorical fields. Comput Geosci 23, 1087–1106 (2019). https://doi.org/10.1007/s10596-019-09874-z
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10596-019-09874-z