The following preprints are independent research by Parke Roberts. Each paper examines whether space weather conditions show a measurable statistical association with specific categories of world events. The research makes no causal claim. No external funding or institutional affiliation is involved.
Each entry below links to a plain-language summary written for a general audience. The full preprint is also linked for readers who want the complete methodology and results.
2026 · Preprint
Two space weather signals replicated across two independent time windows in a global inventory of referendum events spanning 2010 to 2026. The same signals held specifically for referendums while failing under identical conditions for treaties and coups tested in parallel. Effect sizes are small. The subtype concentration and cross-window persistence are what make the finding worth reporting.
Plain-language summary →2026 · Preprint
Broad-spectrum geomagnetic suppression in the 10 days before major earthquakes (M≥5) was statistically enriched relative to 26,250 matched control windows across seven years. The enrichment held across four structurally distinct control designs, all seven leave-one-year-out configurations, and duplicate-date deduplication. A secondary exploratory finding identified a candidate IMF variability transition on earthquake day itself.
Plain-language summary →All findings are independent, unfunded, and not affiliated with any institution. Data sources: NASA OMNI-2, NOAA SWPC, NOAA NCEI, SILSO, University of Oulu neutron monitor.