Authors: Pouya Hosseinzadeh (Utah State University),Soukaina Filali Boubrahimi(Utah State University),Shah Muhammad Hamdi(Utah State University)

We present a multivariate, multi-instrument time-series dataset designed to support the study and prediction of Solar Energetic Particle (SEP) events through consistent characterization of pre-flare solar and heliospheric conditions. The dataset spans 1998–2013 and integrates time-aligned measurements from the OMNI database, including solar-wind plasma parameters and interplanetary magnetic-field components, with GOES soft X-ray and proton flux observations. For each selected solar flare, we extract a uniform 24-hour window preceding the reported flare onset time and assign SEP and non-SEP labels based on cross-referencing the MEMPSEP, GSEP, and NOAA catalogs across four proton energy bands (>10, >30, >60, and >100 MeV). Standardized preprocessing procedures, including temporal alignment, resampling to a five-minute cadence, gap handling, and quality control, ensure consistency across all variables and events. All data are released in harmonized per-variable files accompanied by comprehensive metadata, offering a reproducible and ready-to-use resource for statistical studies, comparative analyses, and the development of space-weather forecasting methods. Future work will extend this resource toward an operational SEP event prediction framework.