Scaling-law-informed neural point processes for earthquake sequence forecasting

Abstract

Earthquake sequence forecasting requires models that can learn nonlinear history dependence while retaining robust statistical structure. We develop a scaling-law-informed neural marked point process, termed Fusion, that combines neural representations of catalog history with temporal features derived from the Epidemic-Type Aftershock Sequence model and magnitude information derived from the Gutenberg--Richter law. The model separates the magnitude cutoff applied to the input catalog from the fixed target-event threshold, allowing lower-magnitude earthquakes to inform forecasts without changing the target-event set. For the 2016--2017 Amatrice--Visso--Norcia sequence, Fusion achieves the highest target-event temporal likelihood when lower-magnitude events are retained, outperforming both ETAS and a purely neural point-process baseline. Event-wise and cumulative analyses show sustained timing gains through substantial portions of the Visso and Norcia sequences. Across five benchmark catalogs, catalog-specific neural training with a fixed ETAS prior yields the highest temporal likelihood at the minimum evaluated magnitude cutoff. Magnitude likelihood shows no consistent predictive gain beyond the Gutenberg--Richter-based ETAS reference, indicating that the additional information captured by Fusion is primarily temporal. These results show that lower-magnitude catalog histories and empirical scaling-law information complement neural sequence learning for target-event timing.

Graph-based probabilistic earthquake clustering and forecasting in Sumatra
Giuseppe Petrillo, Stefania Gentili, Luca Dal Zilio
Tectonophysics1 Oct 2026Peer-reviewedCC BY 4.0Full textGeophysics & seismologyHAL
ANADEF: A Nested-Permutation Alarm for Dual-Parameter Earthquake Forecasting
Hamzeh Mohammadigheymasi, Muhammed Hossein Mousavi, Nuno Pombo
arXiv7 Sep 2026PreprintarXiv licenceFull textGeophysics & seismologyarXiv
Extracting and Characterizing Near-Fault Effects from Strong Motion Observations of the 2022 Chishang Earthquake Sequence
Grigorios Lavrentiadis, Domniki Asimaki, Trevor J. Carey +1
arXiv3 Nov 2026PreprintCC BY 4.0Full textGeophysics & seismologyarXiv
SIETAS model: Swarm-informed triggering, earthquake diffusion, and multiscaling applied to the current Campi Flegrei unrest
Cataldo Godano, Anna Tramelli, Vincenzo Convertito +1
Physical Review Research15 Oct 2026Peer-reviewedCC BY 4.0Full textGeophysics & seismologyHAL
The 2016 Mw 7.0 Kumamoto Earthquake Sequence, Japan revisited: Insights from Spatio-Temporal Analysis of Seismicity Parameters
Hamzeh Mohammadigheymasi, Muhammed Hossein Mousavi, Marjan Tourani +1
arXiv7 Sep 2026PreprintarXiv licenceFull textGeophysics & seismologyarXiv

Cite

BibTeX RIS

BibTeX

@misc{xiong2026scalinglawinformed,
  title = {{Scaling-law-informed neural point processes for earthquake sequence forecasting}},
  author = {Tianlu Xiong and Zaibo Zhao and Yunrui Li and Wenqi Liu and Yosef Ashkenazy and Yongwen Zhang},
  howpublished = {arXiv},
  year = {2026},
  month = nov,
  doi = {10.48550/arxiv.2608.18791},
  eprint = {2608.18791},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2608.18791},
}

RIS

TY  - UNPB
TI  - Scaling-law-informed neural point processes for earthquake sequence forecasting
AU  - Tianlu Xiong
AU  - Zaibo Zhao
AU  - Yunrui Li
AU  - Wenqi Liu
AU  - Yosef Ashkenazy
AU  - Yongwen Zhang
PY  - 2026
DA  - 2026/11/05
PB  - arXiv
DO  - 10.48550/arxiv.2608.18791
UR  - https://arxiv.org/abs/2608.18791
AB  - Earthquake sequence forecasting requires models that can learn nonlinear history dependence while retaining robust statistical structure. We develop a scaling-law-informed neural marked point process, termed Fusion, that combines neural representations of catalog history with temporal features derived from the Epidemic-Type Aftershock Sequence model and magnitude information derived from the Gutenberg--Richter law. The model separates the magnitude cutoff applied to the input catalog from the fixed target-event threshold, allowing lower-magnitude earthquakes to inform forecasts without changing the target-event set. For the 2016--2017 Amatrice--Visso--Norcia sequence, Fusion achieves the highest target-event temporal likelihood when lower-magnitude events are retained, outperforming both ETAS and a purely neural point-process baseline. Event-wise and cumulative analyses show sustained timing gains through substantial portions of the Visso and Norcia sequences. Across five benchmark catalogs, catalog-specific neural training with a fixed ETAS prior yields the highest temporal likelihood at the minimum evaluated magnitude cutoff. Magnitude likelihood shows no consistent predictive gain beyond the Gutenberg--Richter-based ETAS reference, indicating that the additional information captured by Fusion is primarily temporal. These results show that lower-magnitude catalog histories and empirical scaling-law information complement neural sequence learning for target-event timing.
ER  -

Abstract

Related

Cite