Automatic Interface Detection for Electrical Resistivity Tomography Data

Electrical Resistivity Tomography (ERT) can be used for several aspects in mineral, especially sand and gravel deposits in the UK, acquisition. The drawback of the ERT method is primarily its smoothed interfaces of resistivity changes, which is not particular suitable for manually picking. Hence automatic interface detection can be an alternative. The new developed edge detector is based on the validation of Chambers et al. (2010) and Hsu et al. (2010) solutions. It is based on detection of zero-crossings of the second derivative of resistivity data with respect to depth-direction and the classification of the zero-crossings by the first derivative. The effectiveness of the edge detector has been tested on synthetic models. It could be shown that it fulfils the given criteria of investigating all boundaries quite well. The fulfilling of the criteria of the correct location of the investigated boundary is strongly dependent on the inversion process. As well was observed that natural logarithmic data as input data provides with increasing depth better solutions. It become clear further improvements and validations have to be done to improve the edge detector handling and resistant.