Sensitivity analysis of spatial models using geostatistical simulation

Nathalie Saint-Geours (1, 2) , Christian Lavergne (3, 4) , Jean-Stéphane Bailly (1) , Frédéric Grelot (5)

1 UMR TETIS - Territoires, Environnement, Télédétection et Information Spatiale
2 GdR MASCOT-NUM - Méthodes d'Analyse Stochastique des Codes et Traitements Numériques
3 I3M - Institut de Mathématiques et de Modélisation de Montpellier
4 UPVM - Université Paul-Valéry - Montpellier 3
5 UMR G-EAU - Gestion de l'Eau, Acteurs, Usages

Geostatistical simulations are used to perform a global sensitivity analysis on a model Y = f(X1 ... Xk) where one of the model inputs Xi is a continuous 2D-field. Geostatistics allow specifying uncertainty on Xi with a spatial covariance model and generating random realizations of Xi. These random realizations are used to propagate uncertainty through model f and estimate global sensitivity indices. Focusing on variance-based global sensitivity analysis (GSA), we assess in this paper how sensitivity indices vary with covariance parameters (range, sill, nugget). Results give a better understanding on how and when to use geostatistical simulations for sensitivity analysis of spatially distributed models.

Mots clés: Sobol

Mots clés: variance decomposition; geostatistics; Sobol'; sensitivity analysis; spatial model

Domaines: Statistiques [math.ST]; Théorie [stat.TH]

Nathalie Saint-Geours, Christian Lavergne, Jean-Stéphane Bailly, Frédéric Grelot. Sensitivity analysis of spatial models using geostatistical simulation. Mathematical Geosciences at the Crossroads of Theory and Practice - IAMG 2011 Conference, Sep 2011, Salzburg, Austria. pp.178-189, ⟨10.5242/iamg.2011.0172⟩. ⟨hal-00666755⟩