Stochastic Analysis of Saltwater Intrusion in Heterogeneous Aquifers using Local Average Subdivision

Gareth Robinson, Ashraf Ahmed, Gerard Hamill

    Research output: Contribution to conferencePaperpeer-review


    This study investigates the effects of ground heterogeneity, considering permeability as a random variable, on an intruding SW wedge using Monte Carlo simulations. Random permeability fields were generated, using the method of Local Average Subdivision (LAS), based on a lognormal probability density function. The LAS method allows the creation of spatially correlated random fields, generated using coefficients of variation (COV) and horizontal and vertical scales of fluctuation (SOF). The numerical modelling code SUTRA was employed to solve the coupled flow and transport problem. The well-defined 2D dispersive Henry problem was used as the test case for the method. The intruding SW wedge is defined by two key parameters, the toe penetration length (TL) and the width of mixing zone (WMZ). These parameters were compared to the results of a homogeneous case simulated using effective permeability values. The simulation results revealed: (1) an increase in COV resulted in a seaward movement of TL; (2) the WMZ extended with increasing COV; (3) a general increase in horizontal and vertical SOF produced a seaward movement of TL, with the WMZ increasing slightly; (4) as the anisotropic ratio increased the TL intruded further inland and the WMZ reduced in size. The results show that for large values of COV, effective permeability parameters are inadequate at reproducing the effects of heterogeneity on SW intrusion.
    Original languageEnglish
    Number of pages4
    Publication statusPublished - 16 Jun 2014
    EventSWIM23 - 23 Salt Water INtrusion Meeting - Husumhus - 25813 Husum, Germany
    Duration: 16 Jun 201420 Jun 2014


    ConferenceSWIM23 - 23 Salt Water INtrusion Meeting
    CityHusumhus - 25813 Husum


    Dive into the research topics of 'Stochastic Analysis of Saltwater Intrusion in Heterogeneous Aquifers using Local Average Subdivision'. Together they form a unique fingerprint.

    Cite this