GLoBoCat: Global Long-wavelength Body-wave Catalog

GLoBoCat is a dataset for global long-period seismic body wave arrival times, enabled by a machine learning picking model trained on onset arrivals picked by experts (Houser+2008). This work is submitted to Geophysical Research Letters for peer-review. The current catalog comprises ~2 million long period P and S arrivals for earthquakes M>5.5 based on Global CMT and stations available on IRIS Data Management Center between 2010 and 2023. The catalog in ObsPy format obspy_catalog.tar.gz contains hierachical objects Event, Origin, Pick, and Arrival for cross-reference. Loading the complete files may take ten to twenty minutes. We also provide the simplified information focus on arrivals that can be read rapidly with Pandas (pandas_table.tar.gz) or in plain text (text_table.tar.gz), along with the picking model final_model.h5 that can be loaded with EQTransformer (Mousavi+2020).