Authors: Weihao Liu (University of Michigan), Lulu Zhao (University of Michigan), Tamas Gombosi (University of Michigan)

Solar energetic particles (SEPs) are high-energy particles emitted by the Sun during intense solar events, such as solar flares and coronal mass ejections (CMEs), which can pose severe radiation risks to both space-borne assets and astronauts. In order to safeguard our astronauts and valuable instruments in space exploration, predicting the occurrence and properties of energetic particles is essential. When building the prediction model, the quality of the dataset used plays a crucial role in the model development, validation, and evaluation. Although numerous catalogs and models have been developed based on a number of observations from different instruments, we are still lacking a homogeneous dataset that can serve as a benchmark. We propose to construct a comprehensive and homogeneous dataset including SEP events from 1970 and continuing to update.

When building the dataset, we first integrate SEP information from three primary catalogs: the National Oceanic and Atmospheric Administration solar proton events list, the geostationary solar energetic particle catalog (Rotti et al., 2022), and the one maintained by the Geophysical Center of the Russian Academy of Sciences. When integrating the dataset, we encounter discrepancies among these catalogs. We conduct visual examinations, cross-referencing with the plasma and magnetic field properties to determine the correct inputs. 590 events have been identified as SEP events, covering over 5 decades. Based on the dataset, we conduct statistical analysis to study the causality relations between the sources of SEP events and their properties detected at earth. The dataset will facilitate the development of a more comprehensive and robust machine-learning model for the community, enabling accurate predictions of SEPs in the future.