The information in this repository outlines how to extract and process the seismic and infrasound data used to train both seismic and seismoacoustic neural networks as introduced in the manuscript, "Deep Multimodal Learning for Seismoacoustic Fusion to Improve Earthquake-Explosion Discrimination within the Korean Peninsula", (Ronac Giannone et al., 2024, in review). ObsPy and TensorFlow software packages are used for geophysical and deep learning analyses, respectively.
To install: conda env create -f environment.yml
To activate: source activate koreageonet
The data used in this study can be found at 10.5281/zenodo.10795252.
Array locations can be found in Korea_Array_Locations.zip. Earthquake-Explosion databases as well as spreadsheets containing information on individual array detections can be found in Spreadsheets.zip.