A General-Purpose Diversified 2D Seismic Image Dataset from NAMSS

We introduce the Unicamp-NAMSS dataset, a large, diverse, and geographically distributed collection of migrated 2D seismic sections designed to support modern machine learning research in geophysics. We constructed the dataset from the National Archive of Marine Seismic Surveys (NAMSS), which contains decades of publicly available marine seismic data acquired across multiple regions, acquisition conditions, and geological settings. After a comprehensive collection and filtering process, we obtained 2588 cleaned and standardized seismic sections from 122 survey areas, covering a wide range of vertical and horizontal sampling characteristics. To ensure reliable experimentation, we balanced the dataset so that no survey dominates the distribution, and partitioned it into non-overlapping macro-regions for training, validation, and testing. This region-disjoint split allows robust evaluation of generalization to unseen geological and acquisition conditions. We validated the dataset through quantitative and embedding-space analyses using both convolutional and transformer-based models. These analyses showed that Unicamp-NAMSS exhibits substantial variability within and across regions, while maintaining coherent structure across acquisition macro-region and survey types. Comparisons with widely used interpretation datasets (Parihaka and F3 Block) further demonstrated that Unicamp-NAMSS covers a broader portion of the seismic appearance space, making it a strong candidate for machine learning model pretraining. The dataset, therefore, provides a valuable resource for machine learning tasks, including self-supervised representation learning, transfer learning, benchmarking supervised tasks such as super-resolution or attribute prediction, and studying domain adaptation in seismic interpretation.

💡 Research Summary

The paper presents the Unicamp‑NAMSS dataset, a large‑scale, diverse collection of migrated 2‑D seismic sections intended to support modern machine learning research in geophysics. The authors sourced the data from the National Archive of Marine Seismic Surveys (NAMSS), a public repository maintained by the USGS that contains decades‑old marine seismic surveys from various regions, acquisition periods, and geological settings.

Data acquisition began with a manual, map‑based exploration of the NAMSS portal, identifying 547 multichannel 2‑D surveys. Metadata and filename patterns were parsed to locate migrated products, yielding 9 350 migrated files (≈157 GB). Because the volume per survey varied dramatically (from 17 MB to 9 GB), the authors imposed a balancing rule: no single survey contributes more than 300 MB. This threshold reduced the total collected volume to about 33.6 GB, distributed across 3 210 SEG‑Y files.

The preprocessing pipeline extracts raw seismic traces from each SEG‑Y file, converts IBM hexadecimal floating‑point values to IEEE double precision when needed, and assembles the data into matrices of (samples × traces). Each matrix is amplitude‑normalized to the interval