Le jeudi 4 juin 2015 à 10:30, en salle de conf de l'OSUR (campus de Beaulieu, bât.
Le jeudi 4 juin 2015 à 10:30
, en salle de conf de l'OSUR (campus de Beaulieu, bât. 14B, RDC), Arash Massoudieh propose un séminaire intituléUncertainty analysis in groundwater dating with environmental tracers using Markov chain Monte Carlo methodPrésentation :
The estimation of groundwater age has received increasing attention due to its applications in assessing the sustainability of water withdrawal from the aquifers and evaluating the vulnerability of groundwater resources to near surface or recharge water contamination. In most of the works done in the past, whether a single or multiple tracers used for groundwater dating, the uncertainties in observed concentrations of the tracers and their decay rate constants have been neglected. Furthermore, tracers have been assumed to move at the same speed as the groundwater while in reality some retardation might occur in case of some tracers. There are uncertainties on the decay rates of synthetic chemicals such as CFCs commonly used for groundwater dating. All these uncertainties in addition to the noise associated with the measurement can propagate into the estimated values of the Lumped Parameter Model (LPM) parameters. In this presentation development of a Bayesian modeling approach using Markov Chain Monte Carlo method for estimation of age distribution is described. The model considers the uncertainties in the measured tracer concentrations as well as the parameters affecting the concentration of tracers in the groundwater and provides the frequency distributions of the parameters defining the groundwater age distribution. The model also incorporates the effect of the contribution of dissolution of aquifer minerals in diluting the 14C signature and the uncertainties associated with this process on inferred age distribution parameters. The results of application of the method to several case studies including Laselva Biological Station – Costa Rica, as well as some hypothetical cases will also be presented. Some results on the use of temporal tracer data, freeform distributions, and also the use of temporal tracers data to evaluate the temporal variations in flow will also be presented.
Contact :Arash Massoudieh (The Catholic University of America)Contact OSUR :Jean-Raynald de Dreuzy