February 22, 2018
Journal Article

Probability and cumulative density function methods for the stochastic advection-reaction equation

Abstract

We present a cumulative density function (CDF) method for the probabilistic analysis of $d$-dimensional advection-dominated reactive transport in heterogeneous media. We employ a probabilistic approach in which epistemic uncertainty on the spatial heterogeneity of Darcy-scale transport coefficients is modeled in terms of random fields with given correlation structures. Our proposed CDF method employs a modified Large-Eddy-Diffusivity (LED) approach to close and localize the nonlocal equations governing the one-point PDF and CDF of the concentration field, resulting in a $(d + 1)$ dimensional PDE. Compared to the classsical LED localization, the proposed modified LED localization explicitly accounts for the mean-field advective dynamics over the phase space of the PDF and CDF. To illustrate the accuracy of the proposed closure, we apply our CDF method to one-dimensional single-species reactive transport with uncertain, heterogeneous advection velocities and reaction rates modeled as random fields.

Revised: March 19, 2018 | Published: February 22, 2018

Citation

Barajas-Solano D.A., and A.M. Tartakovsky. 2018. Probability and cumulative density function methods for the stochastic advection-reaction equation. SIAM/ASA Journal on Uncertainty Quantification 6, no. 1:180-212. PNNL-SA-123069. doi:10.1137/16M1109163