Severe Weather Suite — May 2024

The whole weather suite on one continental map and one playhead: 19–22 May 2024, the prolonged late-May severe-weather sequence whose 21 May outbreak produced the Greenfield, Iowa EF4. Five independent public NOAA datasets, each built by its own adapter, are streamed together and coordinated by the playback governor.

Bottom to top: HRRR 500 mb steering wind as a dense drifting field of cool-blue particles — thousands of dots gliding with the mid-tropospheric flow the storms ride, so the drift runs with the cells; deliberately low-contrast so it reads as background context beneath the storms rather than competing with them; the MRMS national reflectivity mosaic as moving dBZ isoband polygons — the precipitation field, cross-dissolving between scans — with storm-cell centroids and SCIT cell tracks over it; the WPC surface frontal analysis as the synoptic skeleton above the field, drawn in the full classic notation — spline-smoothed cold, warm, occluded and stationary fronts with their filled triangle and semicircle pips riding the advancing side (derived by matching each front against the next analysis), the stationary ones alternating the classic red/blue with pips on opposite sides, plus dashed troughs — redrawn every three hours and cross-dissolved between analyses; and every GOES-16 GLM lightning flash as an additive splat, flickering individually and stacking into a glow where convection concentrates. All seven archives register as required governor sources, so the shared clock waits for every stream — no layer runs dry and vanishes while the others keep animating — with each widening its own loader window.

It is the payoff of the suite: you watch the wind organize, the fronts march east with the rain bands breaking out along them, the cells and their tracks light up the convective cores, and the lightning flare along the leading edge — cause and effect, one clock. Each feed also plays on its own in the fullscreen viewer: the GLM lightning, the MRMS precipitation field, the HRRR steering wind and the Iowa derecho radar composite.

python glm_lightning.py … && python mrms_weather.py … && python hrrr_advect.py … && python wpc_fronts.py …

Assembles seven archives built by the four weather adapters (goes-glm-lightning, mrms-precip-field/-cells/-tracks, hrrr-wind, wpc-fronts + its -pips companion) into one type:'weather' composite. Rebuild any layer independently.

Full recipes for every dataset live in the data-generation guide.