Mass movements occur frequently in high mountainous regions worldwide, causing severe casualties, economic losses, and persistent threats to ecosystems and infrastructure. In regions characterized by rugged terrain, sparse population, and limited data, early identification and precise monitoring of mass movements remain central challenges. Seismic signals have recently been widely used for mass-movement detection and dynamics inversion due to their capability for continuous and remote monitoring. However, conventional seismic analysis methods remain limited in detecting low-amplitude, low-frequency precursor and initiation signals. To address this, we propose a multi-parameter seismic metric (MSM) that quantifies instantaneous amplitude, short-term energy, and cumulative energy trends, enabling efficient detection and classification of continuous seismic signals from mass movements. Time-frequency analysis of the 2018 Nayong rock avalanche, validated by UAV-based optical-flow measurements, demonstrates that MSM effectively detects and classifies seismic events from fragmented rock collapses, reliably identifying the main avalanche, local failures, and precursor signals. Compared with STA/LTA and Benford’s law, MSM maintains high sensitivity during low-amplitude, low-energy stages. Analysis of the Blatten event shows that MSM effectively detects and classifies ice-rock avalanches, although the composition and integrity of the ice-rock mass influence seismic spectra and energy distribution, reducing sensitivity during ultra-low-frequency initiation. The optimized MSM, combined with Benford’s law, improves detection at this stage. MSM provides a robust and sensitive framework for detecting and classifying main events and precursors of rock and ice-rock avalanches, offering potential support for early warning and risk assessment of mass movements.

Multi-parameter seismic metrics,Seismic signals,Precursor detection,Rock/ice-rock avalanches,High mountainous regions