Time-lapse inversion of self-potential data using particle filter

Abstract The self-potential method is an effective geophysical method for monitoring groundwater and soil pollution. However, due to the flow of groundwater, the observed mixed signals of pollution sources are often constantly changing. How to accurately describe the distribution changes of pollutants is an important issue. We propose an inversion algorithm based on particle filter, which can invert and monitor the movement of underground pollution sources. The pollution source is regarded as a polarized sphere, and then we set up a dynamic numerical model. Then we successfully achieved dynamic monitoring of the burial depth of the model, verifying the effectiveness and accuracy of the algorithm.