I built this index to show where storms overlap the places that relied on Election Day voting in 2024. The saved November 3, 3 p.m. Eastern snapshot combines NOAA radar, short-range forecasts, and NWS alerts. Each populated area receives a hand-set severity score, S, from 0 to 100 for its worst weather through the district's closing window. I allocate historical county and town votes within 2026 districts by population, then weight severity by Election Day votes, E, divided by votes across all methods, T:
Identical weather therefore gets one quarter the score in a district with 20% Election Day voting compared with one at 80%. This measures historical geographic exposure, without estimating lost votes or people yet to vote. Actual method counts are used where available; an EAC participation proxy fills some gaps without creating party counts. Missing observations add no numerator term, while all mapped votes remain in the denominator, and rankings require 95% coverage. The saved explorer and score definitions provide the dated results, coverage, and partisan context.