Groundwater pollution risk using a modified Latin hypercube sampling

Research output: Contribution to journalArticle

3 Citations (Scopus)


Characterisation of groundwater modelling involves significant uncertainty because of estimation errors of these models and other different sources of uncertainty. Deterministic models do not account for uncertainties in model parameters, and thus lead to doubtful output. The main alternatives for deterministic models are the probabilistic models and perturbation methods such as Monte Carlo Simulation (MCS). Unfortunately, these methods have many drawbacks when applied in risk analysis of groundwater pollution. In this paper, a modified Latin Hypercube Sampling method is presented and used for risk, uncertainty, and sensitivity analysis of groundwater pollution. The obtained results were compared with other sampling methods. Results of the proposed method have shown that it can predict the groundwater contamination risk for all values of probability better than other methods, maintaining the accuracy of mean estimation. Sensitivity analysis results reveal that the contaminant concentration is more sensitive to longitudinal dispersivity than to velocity.

Original languageEnglish
Pages (from-to)223-234
Number of pages12
JournalJournal of Hydroinformatics
Issue number3
Publication statusPublished - Jul 2006
Externally publishedYes



  • Contaminant transport modelling
  • Latin hypercube sampling
  • Monte Carlo simulation
  • Probabilistic modelling
  • Risk analysis

ASJC Scopus subject areas

  • Atmospheric Science
  • Geotechnical Engineering and Engineering Geology

Cite this