Combining noisy well data and expert knowledge in a Bayesian calibration of a flow model under uncertainties: an application to solute transport in the Ticino basin

In this talk we present the results of a case-study aimed at providing a UQ analysis of solute travel times in groundwater in the basin of the Ticino River (northern Italy), starting from well data collected over a month in summer 2014. We consider a steady-state groundwater flow model (developed in MODFLOW) and perform a sensitivity analysis using the Morris method to discard uninfluential parameters. We then employ Bayesian inversion (with Gaussian approximation) to obtain a data-informed posterior pdf for the remaining parameters, and propagate these pdfs to travel times computed by particle tracking (MODPATH). The likelihood function employed in the Bayesian inversion takes into account both well measurements and expert knowledge about the extent of the springs in the domain under study.

Contatto:
alessio.fumagalli@polimi.it