Adaptive Multi‐Objective Optimization for Real‐Time Monitoring of Rapid Tracer Transport Using Electrical Resistivity Tomography: Balancing Spatial and Temporal Resolution Article Swipe
YOU?
·
· 2025
· Open Access
·
· DOI: https://doi.org/10.1029/2025wr040444
· OA: W4415388072
Electrical resistivity tomography (ERT) is widely used to monitor electrically conductive tracers. Dense data sets, while improving spatial resolution, often result in long acquisition times, causing temporal smearing of plumes, especially in highly permeable media. In such cases, rapid changes in electrical conductivity may invalidate the assumption of constant subsurface properties during data collection, reducing monitoring accuracy. Previous studies have focused on improving the spatial resolution by optimizing measurement configurations, but a comprehensive analysis of trade‐offs between spatial resolution and measurement time for transport studies remains lacking. This study developed an adaptive multi‐objective ERT survey‐design model to consider both spatial and temporal resolution for monitoring rapidly migrating targets. By adapting the non‐dominated sorting genetic algorithm II, we explored the trade‐offs between these two competing objectives in a synthetic 3‐D aquifer where the conductive tracer migrated approximately 15 m in 4 hr. Results show that ERT data obtained from a specific number of randomly selected and standard measurement configurations yielded limited spatial resolution. Moreover, excessive data collection hindered plume characterization due to the plume's rapid migration during the prolonged survey time. In contrast, the proposed method effectively resolved conflicts between spatial and temporal resolution, providing Pareto‐optimal solutions for time‐lapse ERT surveys. The Pareto front identified optimal combinations of measurement configurations that maximize spatial resolution and minimize data acquisition time, thereby enhancing real‐time monitoring of fast‐migrating plumes. Compared to the standard data set, the estimation accuracy of total solute mass evolution improved by up to 22%.
