Time-Lapse 4-D Resistivity Imaging Inversion with Positivity Constraints
M.H. Loke, P.B. Wilkinson, T. Dahlin, J.E. Chambers, S. Uhlemann and T. Dijkstra
Event name: 24th European Meeting of Environmental and Engineering Geophysics
Session: Modelling, Inversion and Data-processing in Near Surface Geophysics I
Publication date: 09 September 2018
Info: Extended abstract, PDF ( 1.6Mb )
Price: € 20
Time-lapse resistivity surveys are used to monitor changes in the subsurface. In some situations, it is known that the resistivity will only decrease (or increase) with time. The 4-D ERT smoothness-constrained inversion method, that includes temporal smoothness constraint, has proved to be a robust method that reduces artefacts due to noise. However, in some cases, the time-lapse inverse models might show an increase in the resistivity with time where it is only expected to decrease. We modify the 4-D ERT inverse method to include a constraint that removes this artefact. The standard 4-D ERT inversion algorithm is first used to generate an initial model. If the resistivity is expected to decrease with time, for the model cells that show a resistivity increase with time, a truncation procedure is used where the resistivities of the different time models are reset to the mean value (corresponding to zero change with time). We then use the method of transformations in the inversion method that ensures the resistivities of the later time models are always less than the first model. The constraints can be modified so that they are only applied to selected regions in the model in cases where additional information is available.