You can manage bookmarks using lists, please log in to your user account for this.
Media type:
E-Article
Title:
Calibration of seawater intrusion models: Inverse parameter estimation using surface electrical resistivity tomography and borehole data
Contributor:
Beaujean, J.
[Author];
Nguyen, F.
[Author];
Kemna, A.
[Author];
Antonsson, A.
[Author];
Engesgaard, P.
[Author]
imprint:
AGU, 2014
Published in:Water resources research 50(8), 6828 - 6849 (2014). doi:10.1002/2013WR014020
Language:
English
DOI:
https://doi.org/10.1002/2013WR014020
ISSN:
1944-7973;
0148-0227;
0043-1397
Origination:
Footnote:
Diese Datenquelle enthält auch Bestandsnachweise, die nicht zu einem Volltext führen.
Description:
Electrical resistivity tomography (ERT) can be used to constrain seawater intrusion models because of its high sensitivity to total dissolved solid contents (TDS) in groundwater and its relatively high lateral coverage. However, the spatial variability of resolution in electrical imaging may prevent the correct recovery of the desired hydrochemical properties such as salt mass fraction. This paper presents a sequential approach to evaluate the feasibility of identifying hydraulic conductivity and dispersivity in density-dependent flow and transport models from surface ERT-derived mass fraction. In the course of this study, geophysical inversion was performed by using a smoothness constraint Tikhonov approach, whereas the hydrological inversion was performed using a gradient-based Levenberg-Marquardt algorithm. Two synthetic benchmarks were tested. They represent a pumping experiment in a homogeneous and heterogeneous coastal aquifer, respectively. These simulations demonstrated that only the lower salt mass fraction of the seawater-freshwater transition zone can be recovered for different times. This ability has here been quantified in terms of cumulative sensitivity and our study has further demonstrated that the mismatch between the targeted and the recovered salt mass fraction occurs from a certain threshold. We were additionally able to explore the capability of sensitivity-filtered ERT images using ground surface data only to recover (in both synthetic cases) the hydraulic conductivity while the dispersivity is more difficult to estimate. We attribute the latter mainly to the lack of ERT-derived data at depth (where resolution is poorer) as well as to the smoothing effect of the ERT inversion.