Are Seismic Foundation Models More Than Glorified Feature Extractors?

Laura Laurenti; David Liu; Brian Kulis; Men-Andrin Meier; Christopher Johnson

cs.LGOct 27, 2026

Are Seismic Foundation Models More Than Glorified Feature Extractors?

Authors: Laura Laurenti, David Liu, Brian Kulis, Men-Andrin Meier, Christopher Johnson

Organizations: Institute of Geophysics, ETH Zurich · Swiss Seismological Service, ETH Zurich · Dept. of Computer Science, Boston University · Dept. of Electrical and Computer Engineering, Boston University ·

Abstract

Earthquake sequences are stochastic on two separated timescales: the seismic waves of any individual earthquake propagate over seconds to minutes, while the earthquakes themselves arrive as a sparse, self-exciting occurrence process, with inter-event intervals from seconds to months. Respecting both scales in one training objective is the open challenge. Foundation models (FMs) for seismic waveforms instead import their learning targets from audio and vision, where every target is computed from a single fixed-length window. Such a window holds one realization of the wave propagation (an earthquake trace), never the occurrence process that decides when earthquakes happen. We develop nine model designs sharing one Mamba encoder-decoder, trained on a fixed corpus of 3M windows, varying the self-supervised objective, quantization bottleneck and mixer; the two key comparisons are architecture-matched, varying the objective alone. Each checkpoint is evaluated frozen on three downstream tasks. Varying the target redistributes downstream skill rather than raising it, and no checkpoint wins everywhere. Adding masking pressure to our strongest classifier re-specializes it rather than making it a generalist: it gains 6.9 points of few-shot P-wave picking precision and loses 9.6 points of few-shot classification accuracy. By the standard imported from language and vision, these models are feature extractors rather than FMs. Controls on capacity, architecture, and domain proximity do not explain why skill moves rather than rises. Against task-specific baselines the advantage is regime-dependent: our checkpoints beat them few-shot on classification and ground-motion prediction, and trail them on full data. We argue the binding constraint is the target, and that settling the question requires a learning target defined across events, the timescale that makes seismicity stochastic in the first place.

Figures & tables

Objective	FEQ/AEQ (%) $\uparrow$		PGA GCN $\downarrow$		PGA linear $\downarrow$		P-prec (%) $\uparrow$		S-prec (%) $\uparrow$
Objective	full	few	full	few	full	few	full	few	full	few
Task-specific baseline	99.2	82.4	0.20	0.331	—	—	96.5	86.8	93.8	87.7
Rec	87.2	50.6	0.344	0.328	0.338	0.326	93.0	18.8	93.5	66.1
Rec+Mag	77.8	63.8	0.198	0.299	0.213	0.288	89.5	45.3	92.5	74.4
Rec+Disc+Mag	93.9	85.4	0.230	0.306	0.197	0.285	93.9	68.8	93.9	82.9
Rec+Disc+Mag+Mask	87.2	82.9	0.161	0.268	0.162	0.263	95.8	83.5	94.2	88.1

Table 1: Frozen-encoder downstream evaluation, full-data and few-shot (

{\approx}100

labelled examples). FEQ/AEQ : foreshock/aftershock accuracy; PGA : peak-ground-acceleration MSE, reported for both ground-motion heads we train (a GCN over a station-distance graph and an independent-station linear head); P-/S-prec : picking precision at

0.5

s tolerance. Baselines are the external task-specific models; the ground-motion baseline is itself a GCN, so it is matched by our GCN column and has no linear-head counterpart in the literature (dashes). Bold marks the best value in a column, underline our best checkpoint; no ordering survives across tasks.

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Figure A1: The three structural differences between the signals audio and time-series FMs are built for and the sequences seismology cares about. (1) Audio and forecasting corpora sit on a fixed sampling clock, whereas earthquake occurrence is event-driven with inter-event gaps from seconds to months, the reason a window-level target cannot reference the occurrence process. (2) Audio carries roughly stationary energy throughout; seismic traces are a noise floor punctuated by rare, high-energy transients: the asymmetry that makes an $L_{1}/L_{2}$ objective mis-weighted for a heavy-tailed law. (3) Audio codecs can design around a known audible band and discard the rest; seismology has no universally safe band, since picking, magnitude, and engineering applications each depend on a different part of the spectrum, so any compression target pre-selects which downstream tasks survive it.

Time-Series Foundation Models

Time-series foundation models are pretrained sequence models designed to transfer across forecasting, classification, anomaly detection, and other temporal tasks.

See all 304 papers

Jun 14, 2026

Domain-Guided Prompting of the Segment Anything Model for Seismic Interpretation: The Role of Attributes, Visualization, and Hybrid Prompts

The advent of large pretrained foundation models for computer vision has significantly improved the efficiency of visual data interpretation. The Segment Anything Model (SAM), in particular, offers powerful zero shot segmentation capabilities through prompt based interaction, thus making it a promising tool for seismic interpretation. However, most existing applications of SAM rely on fine tuning for specific geological targets, which requires extensive labeled data, incurs high computational cost, and often compromises the model's generalization capability. In this study, we introduce a principled framework for zero shot adaptation of foundation models to seismic data. The framework is built on two key components: (1) aligning seismic attributes and visualization choices (e.g., colormaps) with the geological target of interest, and (2) employing a hybrid prompting strategy that combines sparse user defined point prompts with dense mask prompts derived from SAM's internal feature activations. We systematically evaluate this framework across multiple geological targets, datasets, prompt configurations, and seismic attribute representations. Our results demonstrate that geologic target aware selection of seismic attributes and colormaps, combined with hybrid prompting, enhances the separability of geological features and improves boundary delineation and segmentation accuracy relative to point based prompting alone. Our findings show that, when these components are jointly applied, SAM can achieve competitive segmentation performance in a fully zero shot setting, thereby eliminating the need to retrain SAM for each geologic feature. This work establishes a practical and scalable pathway to leverage foundation models in seismic interpretation, reducing reliance on labeled data while preserving model generality.

Aniq Ahmad, Heather Bedle, Ahmad Mustafa

School of Geosciences, University of Oklahoma, Norman, Oklahoma, USA · King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia

Promptable Image Segmentation Vision Foundation Model Adaptation Zero-Shot Referring Image Segmentation Semantic Segmentation Prompt Engineering

Ask PDF

Apr 30, 2026

Parameter-Efficient Adaptation of Pre-Trained Vision Foundation Models for Active and Passive Seismic Data Denoising

The demand for high-resolution subsurface imaging and continuous Earth monitoring has driven rapid growth in active and passive seismic data from dense geophone deployments, distributed acoustic sensing (DAS) arrays, and large-scale 2D and 3D surveys. This expansion makes complex noise suppression increasingly challenging, especially when signal fidelity must be preserved. Conventional supervised deep learning methods are often task-specific, require large paired datasets, and can suffer from domain shift under new acquisition conditions. Foundation models offer a promising alternative, but pre-training seismic foundation models from scratch requires massive domain-specific data and substantial computation. We propose an efficient framework that repurposes general-purpose Vision Foundation Models (VFMs) for geophysical tasks through Parameter-Efficient Fine-Tuning. The architecture uses a pre-trained VFM, a DINOv3 encoder, adapted with Low-Rank Adaptation (LoRA) to enable effective feature adaptation with few additional parameters. To improve robustness under unseen field conditions without ground truth, we introduce a kurtosis-guided unsupervised test-time adaptation module that updates only LoRA parameters during inference. This module self-calibrates the model to site-specific noise by identifying information-rich regions via kurtosis and performing self-training without labeled data. Experiments on public exploration seismic images and DAS vertical seismic profiling data from the Utah FORGE site show that the framework matches or outperforms domain-specific models. Tests on unseen cross-site data from a land survey in China and the Groß Schönebeck geothermal site in Germany further demonstrate strong generalization and effective signal-noise separation. These results highlight the potential of adapting pre-trained VFMs to data-intensive problems in exploration seismology.

Jiahua Zhao, Umair bin Waheed, Jing Sun +3

Computation-based Science and Technology Research Center, The Cyprus Institute, Nicosia, Cyprus · Department of Geosciences, College of Petroleum Engineering and Geosciences, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia · Department of Intelligent Systems, Faculty of Electrical Engineering, Mathematics and Computer Science, Delft University of Technology, Delft, Netherlands +2

Test-Time Adaptation Vision Foundation Model Adaptation Seismology Parameter-Efficient Fine-Tuning

Ask PDF

Sep 26, 2026

Reuse or Relearn? A Spectral View of Earth Observation Foundation Models

Foundation models are rarely used as generic, frozen feature extractors; instead, they are fine-tuned for the target downstream application. This practice is particularly prevalent in Earth observation (EO), and it raises a question that downstream accuracy alone cannot answer: does fine-tuning reuse the pretrained representation, or does it relearn a new one? We study this with spectral diagnostics that compare a model before and after adaptation, quantifying how well its dominant singular subspaces are preserved, how broadly the weight update is distributed, and how large it is. Using natural image models such as CLIP and DINO as a reference, we find that, under the evaluated fine-tuning settings, EO models undergo far larger, higher-rank updates and retain much less of their pretrained structure, so their downstream performance is often obtained with substantial changes to the pretrained weight structure. The diagnostics further provide insight into how cheaply a model can be adapted: where the pretrained subspaces are preserved, adapting a small fraction of the parameters can match full fine-tuning, and where they are not, it can fall behind. More broadly, foundation models, and EO foundation models in particular, should be assessed not only by benchmark accuracy, but also by how reusable their pretrained representation is.

Mehmet Ozgur Turkoglu, Valerio Marsocci, Dominik J. Mühlematter +3

Agroscope, EO of Agroecosystems · ESA, Φ-lab · ETH Zurich

Fine-Tuning Geospatial Foundation Models Representation Drift Parameter-Efficient Fine-Tuning

Ask PDF

Abstract

Time-Series Foundation Models

Explore similar work

Domain-Guided Prompting of the Segment Anything Model for Seismic Interpretation: The Role of Attributes, Visualization, and Hybrid Prompts

Parameter-Efficient Adaptation of Pre-Trained Vision Foundation Models for Active and Passive Seismic Data Denoising

Reuse or Relearn? A Spectral View of Earth Observation Foundation Models