---

Discriminating Landslide Waveforms in Continuous Seismic Data Using Power Spectral Density Analysis.

A global spectral models have been developed for discriminating landslide seismic signals from earthquakes and noise. Landslides exhibit distinct power decay patterns and characteristic power distribution shapes, enabling effective identification. The method was statistically validated using 835 seismic recordings of global landslides. Results highlight the robustness of frequency-power patterns for detecting landslides in continuous seismic data for real-time monitoring and early warning systems to mitigate cascading disasters effectively. The proposed approach is simple, computationally less intensive, statistically significant, and promising for real‐time detection of landslides for developing early warning systems based on seismic networks in the future.

---