Full text loading...
-
Stochastic Time-Lapse Ert Inversion by Means of an Ensemble-Based Approach
- Publisher: European Association of Geoscientists & Engineers
- Source: Conference Proceedings, NSG2022 28th European Meeting of Environmental and Engineering Geophysics, Sep 2022, Volume 2022, p.1 - 5
Abstract
Summary
Time-Lapse Electrical Resistivity Tomography (TL-ERT) is a time-extension of the more conventional ERT method widely used in mapping near-surface processes. The ERT inverse problem is ill-posed and it is usually solved by gradient-based least square approach without a robust uncertainty appraisals. In this work, we modify the stochastic Ensemble Based (EB) algorithm to perform a Time-Lapse ERT inversion simultaneously retrieving the initial resistivity, the resistivity variation in time, and also the corresponding uncertainties.
Specifically, the EB algorithm is an iterative data assimilation method that updates an ensemble of prior realizations based on the difference between observed an predicted data. The final result is an ensemble of posterior realizations from which the mean model and the standard deviation is estimated.
To verify the reliability of the Time-Lapse EB approach, we apply the algorithm to synthetic TL-ERT data simulating an advancing subsurface plume. Then, we use the same inversion strategy for data acquired by Pillemark landfill monitoring station (Samsø, Denmark) and the estimated models are compared with those obtained by a classic deterministic approach. From the inversion results, it emerges that the Time-Lapse EB algorithm is able to correctly detect the subsurface resistivity variations and to estimate the related uncertainties.
© EAGE Publications BV