Multimodel Bayesian analysis of groundwater data worth

Liang Xue, Dongxiao Zhang, Alberto Guadagnini, Shlomo P Neuman

Research output: Contribution to journalArticle

18 Citations (Scopus)

Abstract

We explore the way in which uncertain descriptions of aquifer heterogeneity and groundwater flow impact one's ability to assess the worth of collecting additional data. We do so on the basis of Maximum Likelihood Bayesian Model Averaging (MLBMA) by accounting jointly for uncertainties in geostatistical and flow model structures and parameter (hydraulic conductivity) as well as system state (hydraulic head) estimates, given uncertain measurements of one or both variables. Previous description of our approach was limited to geostatistical models based solely on hydraulic conductivity data. Here we implement the approach on a synthetic example of steady state flow in a two-dimensional random log hydraulic conductivity field with and without recharge by embedding an inverse stochastic moment solution of groundwater flow in MLBMA. A moment-equations-based geostatistical inversion method is utilized to circumvent the need for computationally expensive numerical Monte Carlo simulations. The approach is compatible with either deterministic or stochastic flow models and consistent with modern statistical methods of parameter estimation, admitting but not requiring prior information about the parameters. It allows but does not require approximating lead predictive statistical moments of system states by linearization while updating model posterior probabilities and parameter estimates on the basis of potential new data both before and after such data are actually collected. Key Points Joint consideration of geostatistical and flow model uncertainties Combined assessment of added hydraulic conductivity and head data worth Inverse solution of stochastic moment equations combined with MLBMA

Original languageEnglish (US)
Pages (from-to)8481-8496
Number of pages16
JournalWater Resources Research
Volume50
Issue number11
DOIs
StatePublished - 2014

Fingerprint

Bayesian analysis
groundwater
hydraulic conductivity
hydraulic head
groundwater flow
recharge
aquifer

Keywords

  • data worth
  • inverse stochastic moment solution
  • maximum likelihood Bayesian model averaging
  • model uncertainty

ASJC Scopus subject areas

  • Water Science and Technology

Cite this

Multimodel Bayesian analysis of groundwater data worth. / Xue, Liang; Zhang, Dongxiao; Guadagnini, Alberto; Neuman, Shlomo P.

In: Water Resources Research, Vol. 50, No. 11, 2014, p. 8481-8496.

Research output: Contribution to journalArticle

Xue, Liang ; Zhang, Dongxiao ; Guadagnini, Alberto ; Neuman, Shlomo P. / Multimodel Bayesian analysis of groundwater data worth. In: Water Resources Research. 2014 ; Vol. 50, No. 11. pp. 8481-8496.
@article{3fdefdb317274dadab14dd376ae32b1b,
title = "Multimodel Bayesian analysis of groundwater data worth",
abstract = "We explore the way in which uncertain descriptions of aquifer heterogeneity and groundwater flow impact one's ability to assess the worth of collecting additional data. We do so on the basis of Maximum Likelihood Bayesian Model Averaging (MLBMA) by accounting jointly for uncertainties in geostatistical and flow model structures and parameter (hydraulic conductivity) as well as system state (hydraulic head) estimates, given uncertain measurements of one or both variables. Previous description of our approach was limited to geostatistical models based solely on hydraulic conductivity data. Here we implement the approach on a synthetic example of steady state flow in a two-dimensional random log hydraulic conductivity field with and without recharge by embedding an inverse stochastic moment solution of groundwater flow in MLBMA. A moment-equations-based geostatistical inversion method is utilized to circumvent the need for computationally expensive numerical Monte Carlo simulations. The approach is compatible with either deterministic or stochastic flow models and consistent with modern statistical methods of parameter estimation, admitting but not requiring prior information about the parameters. It allows but does not require approximating lead predictive statistical moments of system states by linearization while updating model posterior probabilities and parameter estimates on the basis of potential new data both before and after such data are actually collected. Key Points Joint consideration of geostatistical and flow model uncertainties Combined assessment of added hydraulic conductivity and head data worth Inverse solution of stochastic moment equations combined with MLBMA",
keywords = "data worth, inverse stochastic moment solution, maximum likelihood Bayesian model averaging, model uncertainty",
author = "Liang Xue and Dongxiao Zhang and Alberto Guadagnini and Neuman, {Shlomo P}",
year = "2014",
doi = "10.1002/2014WR015503",
language = "English (US)",
volume = "50",
pages = "8481--8496",
journal = "Water Resources Research",
issn = "0043-1397",
publisher = "American Geophysical Union",
number = "11",

}

TY - JOUR

T1 - Multimodel Bayesian analysis of groundwater data worth

AU - Xue, Liang

AU - Zhang, Dongxiao

AU - Guadagnini, Alberto

AU - Neuman, Shlomo P

PY - 2014

Y1 - 2014

N2 - We explore the way in which uncertain descriptions of aquifer heterogeneity and groundwater flow impact one's ability to assess the worth of collecting additional data. We do so on the basis of Maximum Likelihood Bayesian Model Averaging (MLBMA) by accounting jointly for uncertainties in geostatistical and flow model structures and parameter (hydraulic conductivity) as well as system state (hydraulic head) estimates, given uncertain measurements of one or both variables. Previous description of our approach was limited to geostatistical models based solely on hydraulic conductivity data. Here we implement the approach on a synthetic example of steady state flow in a two-dimensional random log hydraulic conductivity field with and without recharge by embedding an inverse stochastic moment solution of groundwater flow in MLBMA. A moment-equations-based geostatistical inversion method is utilized to circumvent the need for computationally expensive numerical Monte Carlo simulations. The approach is compatible with either deterministic or stochastic flow models and consistent with modern statistical methods of parameter estimation, admitting but not requiring prior information about the parameters. It allows but does not require approximating lead predictive statistical moments of system states by linearization while updating model posterior probabilities and parameter estimates on the basis of potential new data both before and after such data are actually collected. Key Points Joint consideration of geostatistical and flow model uncertainties Combined assessment of added hydraulic conductivity and head data worth Inverse solution of stochastic moment equations combined with MLBMA

AB - We explore the way in which uncertain descriptions of aquifer heterogeneity and groundwater flow impact one's ability to assess the worth of collecting additional data. We do so on the basis of Maximum Likelihood Bayesian Model Averaging (MLBMA) by accounting jointly for uncertainties in geostatistical and flow model structures and parameter (hydraulic conductivity) as well as system state (hydraulic head) estimates, given uncertain measurements of one or both variables. Previous description of our approach was limited to geostatistical models based solely on hydraulic conductivity data. Here we implement the approach on a synthetic example of steady state flow in a two-dimensional random log hydraulic conductivity field with and without recharge by embedding an inverse stochastic moment solution of groundwater flow in MLBMA. A moment-equations-based geostatistical inversion method is utilized to circumvent the need for computationally expensive numerical Monte Carlo simulations. The approach is compatible with either deterministic or stochastic flow models and consistent with modern statistical methods of parameter estimation, admitting but not requiring prior information about the parameters. It allows but does not require approximating lead predictive statistical moments of system states by linearization while updating model posterior probabilities and parameter estimates on the basis of potential new data both before and after such data are actually collected. Key Points Joint consideration of geostatistical and flow model uncertainties Combined assessment of added hydraulic conductivity and head data worth Inverse solution of stochastic moment equations combined with MLBMA

KW - data worth

KW - inverse stochastic moment solution

KW - maximum likelihood Bayesian model averaging

KW - model uncertainty

UR - http://www.scopus.com/inward/record.url?scp=84919681031&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84919681031&partnerID=8YFLogxK

U2 - 10.1002/2014WR015503

DO - 10.1002/2014WR015503

M3 - Article

AN - SCOPUS:84919681031

VL - 50

SP - 8481

EP - 8496

JO - Water Resources Research

JF - Water Resources Research

SN - 0043-1397

IS - 11

ER -