Groundwater Flow Monitoring by Fusion Probability Tomography of Self-Potential Data

Groundwater flow could produce an observable self-potential (SP) signal; it is, therefore, useful to use SP signal to investigate the condition of water flow, which is meaningful to the processes to which the water flow is critical, such as landslide warning, contaminant transport, and volcanic eruption. Most of the SP data inversion methods are based on prior in situ electrical resistivity information, which is difficult to be accurately obtained in the continuous monitoring of the vadose zone. We suggest a fusion scheme of integrating the electric charge occurrence probability (ECOP) tomography and the continuous complex wavelet transform (CCWT) method to process and interpret SP data when it lacks prior electrical structure information on the background media. A model of fracture flow with precipitation recharge is used to testify our proposed method. The results indicate that the combination scheme can locate the source of SP anomalies with improved accuracy. The proposed fusion probability tomography can locate the fracture flow with an explicit region of probability anomaly and suppress the irrelevant noises.