Tidings
Back to latest
ScienceUnited States

Researchers identify forecasting models that predict earthquake patterns with greater accuracy

A new study reveals certain earthquake forecasting models can reliably anticipate seismic activity patterns for the following day, offering a potential tool for risk assessment in vulnerable regions.

AI-written
Source/event date: 15 Oct 2026

Written automatically from the sources below and not individually reviewed by a human editor.

AI narration · generated on first listen
AI-generated editorial image — not evidence and does not depict the actual event.

Scientists have identified specific earthquake forecasting models capable of consistently describing where, when, and how seismic activity may unfold over the next day. The findings stem from a decade-long evaluation of multiple models, which tested their ability to match real-world observations in California. While earthquakes remain unpredictable in absolute terms, these models now provide a structured approach to assessing risk, potentially aiding those living in high-threat areas. The research team has also released an open-source software package, allowing others to replicate and build upon the analysis. This development could support government agencies, emergency responders, and communities in making more informed decisions about preparedness and response efforts. The study was supported by international funding bodies, including organizations focused on geological research and disaster mitigation. By refining forecasting techniques, the work may contribute to broader efforts in seismic hazard assessment worldwide, particularly in regions prone to frequent tremors. The open-access nature of the tools ensures transparency and collaboration across the scientific community.

How this was made

Generated from retrieved source material, checked for evidence and provenance, and released through runtime safety controls.

Read our editorial standard →

Spot a problem?

Reports go to the operator review queue and do not alter the article automatically.

More in Science