The baseline RGT predictions of 3,000 field datasets for "Pretrain-to-alignment learning paradigm for improving the geophysical AI applicability: A case study in RGT estimation"

In this study, we propose a geological-prior-driven pretrain-to-alignment learning framework for relative geologic time (RGT) estimation in seismic clinothems. The proposed framework systematically integrates self-supervised pretraining, synthetic supervision, geological prior-driven refinement, and domain-adaptation fine-tuning within a unified progressive learning strategy. The final baseline RGT predictions (Field2d_final_baseline), the predictions after synthetic supervison (Field2d_predictions_step1), the predictions after prior-driven refinement, for 3,000 dievrse field datasets (DATA). These baseline results are created by the Computational Interpretation Group (CIG) for for automatic relative geologic time (RGT) estimation, Hui Gao, Xinming Wu, Jiarun Yang, Zhixiang Guo, and Yimin Dou are the main contributors. The source codes for "Pretrain-to-alignment learning paradigm for improving the geophysical AI applicability: A case study in RGT estimation" are publicly available at GitHub (Source_Code).