Data and code for prediction-guided adaptive temporal sampling of time-lapse electrical resistivity tomography

<p dir="ltr">This record contains the synthetic datasets and trajectory-level partitions, processed numerical and sandbox reference trajectories, frozen model components, analysis code, and tabulated prediction and scheduling results accompanying the study of prediction-guided adaptive temporal sampling for time-lapse electrical resistivity tomography. Equal-Budget Fixed, OBS-AS, Linear-AS, and LSTM-AS share 12 initialization samples.

The main acquisition budgets are 42 for the synthetic and numerical evaluations and 20 for the sandbox evaluation. The sandbox data comprise 1,193 interpolated reference values on a nominal 1-min grid derived from inversion-based ROI indicators; they do not represent 1,193 independent TLERT acquisitions. Raw apparent-resistivity measurements, complete inversion projects, and the original sandbox interpolation-generation script are not included.