Processed seismic parameters and event catalogs for machine-learning eruption forecasting at Ubinas volcano, Peru

<p dir="ltr">This dataset contains the processed minute-resolution seismic parameter datasets and event catalogs used to develop and evaluate a multi-station machine-learning system for short-term forecasting of explosive volcanic activity at Ubinas volcano, Peru.</p><p dir="ltr">The archive includes four processed seismic-parameter datasets derived from the 2019 and 2023 eruptive crises, together with event catalogs containing 9 events from 2019 and 76 events from 2023.

The data include Shannon entropy, kurtosis, frequency index, and seismic spectral amplitude measurement (SSAM), computed from the seismic stations analyzed in the study.</p><p dir="ltr">These materials support the manuscript entitled “A transferable multi-station machine-learning system for short-term eruption forecasting: application to Ubinas volcano (Peru).” They are intended to reproduce the data-processing, training, validation, and figure-generation workflows provided in the companion software repository: github.com/UGRRiky/ubinas-ml-eruption-forecasting.</p><p dir="ltr">The complete continuous seismic waveform archive is not included because of its volume and institutional data-sharing restrictions.