SeisMoLLM: Advancing Seismic Monitoring via Cross-modal Transfer with Pre-trained Large Language Model

2025-02-27

AI Summary

To address signal degradation, scarce labeled data, and poor generalization across multiple tasks in earthquake monitoring, this paper proposes the first cross-modal transfer learning paradigm—adapting a pretrained language model (GPT-2) to seismic waveform analysis *without* earthquake-specific pretraining. Our method introduces: (1) a learnable seismic waveform tokenization scheme that maps time-series waveforms to discrete tokens; (2) a cross-modal representation alignment and multi-task joint optimization framework; and (3) unified modeling of five core seismic tasks on the DiTing and STEAD benchmarks. Experiments demonstrate state-of-the-art performance on 36 out of 43 evaluation metrics, including 12 out of 16 few-shot metrics, with relative improvements predominantly ranging from 10% to 50%. Inference latency matches that of lightweight models. This work establishes a scalable, cross-modal foundation model paradigm for seismology.

Technology Category

• Machine Learning: Multimodal Learning
• Natural Language Processing: Language Grounding & Multi-modal NLP
• Computer Vision: Multi-modal Vision

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Recent advances in deep learning have revolutionized seismic monitoring, yet developing a foundation model that performs well across multiple complex tasks remains challenging, particularly when dealing with degraded signals or data scarcity. This work presents SeisMoLLM, the first foundation model that utilizes cross-modal transfer for seismic monitoring, to unleash the power of large-scale pre-training from a large language model without requiring direct pre-training on seismic datasets. Through elaborate waveform tokenization and fine-tuning of pre-trained GPT-2 model, SeisMoLLM achieves state-of-the-art performance on the DiTing and STEAD datasets across five critical tasks: back-azimuth estimation, epicentral distance estimation, magnitude estimation, phase picking, and first-motion polarity classification. It attains 36 best results out of 43 task metrics and 12 top scores out of 16 few-shot generalization metrics, with many relative improvements ranging from 10% to 50%. In addition to its superior performance, SeisMoLLM maintains efficiency comparable to or even better than lightweight models in both training and inference. These findings establish SeisMoLLM as a promising foundation model for practical seismic monitoring and highlight cross-modal transfer as an exciting new direction for earthquake studies, showcasing the potential of advanced deep learning techniques to propel seismology research forward.

Problem

• Cross-modal transfer for seismic monitoring
• Utilizing pre-trained large language models
• Achieving state-of-the-art seismic task performance

Innovation

• Cross-modal transfer technique
• Pre-trained GPT-2 model
• Waveform tokenization method