Monitoring Leachate Dispersion in Landfills using Electrical Resistivity Tomography Inversion based on Deep Learning and Data Assimilation

Groundwater leachate dispersion in landfills is a critical environmental monitoring task for effective landfill management.. Electrical Resistivity Tomography (ERT), a noninvasive geophysical exploration technique, is widely used in groundwater environmental monitoring and landfill management. It measures the distribution of electrical resistivity in subsurface media, offering an effective means to study leachate dispersion processes in landfills.This paper proposes an innovative approach that combines static reconstruction with dynamic assimilation using the Unscented Kalman Filter (UKF) for monitoring leachate dispersion in landfills. Static resistivity models are obtained through a deep learning-based prediction model, and the UKF algorithm is employed for Unscented Kalman Filter assimilation of time-series data to achieve dynamic predictions of resistivity model changes. By comparing with the traditional inversion method, it is shown that the method proposed in this paper can clearly invert the change of resistivity, can portray the size and diffusion direction of the pollutant plume, and realize the dynamic monitoring of the landfill leachate diffusion.