How accurate are the resistivity values from the electrical resistivity tomography (ERT)?

Modern ERT methods/devices can provide 2D resistivity profiles for some terrain. How accurate are the resistivity values in those profiles?

Are these some sorts of apparent resistivities and one should interpret them in a special way or the profile pictures show pretty exact resistivity values in all points?

And if they are not exact then what can cause inaccuracies: buried wires (or other metallic structures), voids etc.?

Maybe some methods/devices are less exact than others?

Inaccuracies are inherent based on a wide range of variables. In particular, the array type, as well as the inversion will affect the resistivity of any layer or body.

Interpreting data based purely on apparent resistivity isn't a good idea at all! You should always try to invert the data to produce realistic conductivity distributions.

The inaccuracies are caused by the path of the current through the Subsurface. This is why the array will affect the final result (different configurations drive current in different ways). Something unusually conductive will divert current towards the body, while resistive bodies will divertv away.

To better estimate the Subsurface resistivity, you can constrain the inversion with prior knowledge. For example, sharp high contrast boundaries, which can affect the recovered values quite significantly, can be incorporated into most inversion algorithms to (usually) enhance the accuracy of the result.

To answer your question, the inaccuracy is a product of the physics - instrumentation can only do so much before the physics kicks in.

Hope this helps! (Please correct me if/where I'm wrong too!)

• Thank you. I found an article that gives some sense of the accuracy of resistivity profiles (with inversion). lup.lub.lu.se/search/record/339568 In the article, there is a comparison of predefined model vs results from inversion technique applied to the apparent resistivity data that calculated again from the predefined model. As far as I can understand this type of verification is quite common for inversion methods. I wonder if there is a comparison of ERT measured resistivities (with inversion) vs those measured some other way (resistivity log? I know very little about this). Commented Oct 13, 2016 at 22:20
• You're quite right, it is common (I'm actually doing that right now!). You should be able to compare the results with borehole resistivity logs too, you're also right. Commented Oct 14, 2016 at 1:16