The metropolis-hastings algorithm, a handy tool for the practice of environmental model estimation : illustration with the biochemical oxygen demand data
Torre, F.; Boreux, Jean-Jacques; Parent, E.
2001 • In Cybergeo: Revue Européenne de Géographie, 187
Parameter uncertainty; Markov Chain Monte Carlo sampling; Bayesian inference
Abstract :
[en] Environmental scientists often face situations where: (i) stimulus-response relationships are non-linear; (ii) data are rare or imprecise; (iii) facts are uncertain and stimulus-responses relationships are questionable.In this paper, we focus on the first two points. A powerful and easy-to-use statistical method, the Metropolis-Hastings algorithm, allows the quantification of the uncertainty attached to any model response. This stochastic simulation technique is able to reproduce the statistical joint distribution of the whole parameter set of any model. The Metropolis-Hastings algorithm is described and illustrated on a typical environmental model: the biochemical oxygen demand (BOD). The aim is to provide a helpful guideline for further, and ultimately more complex, models. As a first illustration, the MH-method is also applied to a simple regression example to demonstrate to the practitioner the ability of the algorithm to produce valid results.
Disciplines :
Mathematics
Author, co-author :
Torre, F.
Boreux, Jean-Jacques ; Université de Liège - ULiège > Département des sciences et gestion de l'environnement > Surveillance de l'environnement
Parent, E.
Language :
English
Title :
The metropolis-hastings algorithm, a handy tool for the practice of environmental model estimation : illustration with the biochemical oxygen demand data
Bates, D. M. and Watts, D. G. (1988). "Nonlinear Regression Analysis and Its Applications". New York: John Wiley & Sons.
Bayes, T. (1763). "An essay towards solving a problem in the doctrine of chances". Phil. Trans. Roy. Soc 53. 370-418.
Reprinted, with an introduction by georges Barnard, in 1958 in Biometrika, 45:293-315.
Berger, J. O. (1985). "Statistical Decision Theory and Bayesian Analysis". New York: Springer Verlag.
Berthouex, P. M. and Brown, L. C. (1994). "Designing experiments to estimate parameters in non linear models". In Statistics for environmental engineers, eds. P. M. Berthouex and L. C. Brown. Boca Raton, USA: Lewis Publishers.
Box, G. E. P. and C., T. G. (1973). "Bayesian inference in statistical analysis". Reading, Massachusetts, USA: Addison-Wesley.
Brooks, S. P. (1998). " Markov chain Monte Carlo and its application,". The Statistician, Journal of the Royal Statistical Society, Series D. 47.1: 69-100.
Carlin, B. P. and Louis, T. A. (1996). "Bayes and empirical Bayes methods for data analysis". London, England: Chapmann & Hall.
Casella, G. and Georges, E. I. (1992). "Explaining the Gibbs Sampler". The American Statistician 46.3: 167-174.
Chaouche A., Parent E.(1999). "Inférence et validation bayésiennes d'un modèle de pluies journalières en régime de mousson". Hydrological Sciences Journal 44(2):199-220. 29
Chen, S., et al. (1990). "Nonlinear system identification using neural networks". Int. J. Control. 51.6: 1215-1228.
DeFinetti, B. (1937). "La prévision, ses lois logiques, ses sources subjectives". Paris: Herman.
Gamerman, D. (1997). "Markov Chain Monte Carlo - Stochastic Simulation for Bayesian inference": Chapman & Hall Eds.
Gelman, A., et al. (1995). "Bayesian Data Analysis". London, UK: Chapman & Hall Eds.
Geman, S. and Geman, D. (1984). "Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images". IEEE Transactions on Pattern Analysis and Machine Intelligence 6. 721-741.
Gilks, W. R., et al., Eds. (1996). "Practical Markov Chain Monte Carlo". New York., Chapman-Hall.
Gilks, W. R., et al. (1992). "Software for the Gibbs sampler". Computing Science and Statistics, 24. 439-448.
Guiot, J., Torre, F., Jolly D., Peyron O., Boreux J.-J. and Cheddadi R. (2000). "Inverse vegetation modelling by Monte Carlo sampling to reconstruct palaeoclimates under changed precipitation seasonality and CO 2 conditions: application to glacial climate in Mediterranean region". Ecological modelling, 127, 119-140.
Hastings, W. K. (1970). "Monte Carlo sampling methods using Markov chains and their application". Biometrika 57. 97-109.
Kiely, G. (1997). "Environmental engineering". Berkshire, England: Irvin / McGraw-Hill.
Krenkel, P. A. and Novotny, V. (1980). "Water quality monitoring". London, UK: Academic Press.
Kuczera, G. (1983). " Improved parameter inference in catchment models: 1. Evaluating parameter uncertainly". Water Resources Research 19.5: 1151-1162.
Kuczera G., Parent E. (1998). "Monte Carlo assessment of parameter uncertainty in conceptual catchment models: the Metropolis algorithm", Journal of. Hydrology (211)1-4:. 69-85.
Lecoutre, B. (1997). "C'est bon à savoir! Et si vous étiez un bayésien qui s'ignore!". Modulad 18. 81-87.
Lindley, D. V. (1972). "Bayesian statistics. A review". CBMS-NSF Regional Conference Series in Applied Mathematics 2. SIAM (Eds). v + 83 pp.
Marske, D. and Polkovski, B. (1972). "Evaluation of methods for estimating biochemical oxygen demand parameters". Journal of Water Pollution Control Federation 44.10: 1987-2000.
Metropolis N., Rosenbluth A.W., Rosenbluth M.N., Teller A.H. and Teller E. (1953). "Equations of state calculations by fast computing machines". J. Chem. Phys. 21. 1087-1092.
Ritter, C. and Tanner, M. A. (1992). "Facilitating the Gibbs Sampler : The Gibbs Stopper and the Griddy-Gibbs Sampler". Journal of the American Statistical Association 87(419): 861-868.
Robert, C. (1992). "L'analyse statistique bayésienne". Paris, France: Economica.
Robert, C. (1996). "Méthodes de Monte-Carlo par chaînes de Markov". Paris, France: Economica.
Tanner, M. H. (1992). "Tools for statistical inference: Observed data and data augmentation methods": Springer-Verlag. 31
Van Straten, G. and Keesman, K. J. (1991). "Uncertainty propagation and speculation in projective forecasts of environmental change: A lake-eutrophication example". J. Forecasting 10. 163-190.