WeatherNext 3 details: real-time satellite training, hourly 5-km forecasts, 60% CRPS gain vs IMERG
ymatias · x · 2026-09-04
A Google team member detailed WeatherNext 3's technical specifics: the model learns directly from global geostationary satellite data, bypassing the typical 6-hour lag of numerical weather prediction to produce hourly forecasts grounded in the latest observations at up to 5-km resolution — critical for fast-developing severe weather. Trained on NASA's IMERG and Google's satellite-radar-based global precipitation reanalysis, it achieves up to 60% CRPS improvement against IMERG, with hourly forecasts at multiple resolutions.
Related event: Google Launches WeatherNext 3, Its Most Advanced Global Weather AI Model(19 posts)→
More from Research
- Stanford team launches ScholarCatalyst benchmark for inspiration retrieval — lateinteraction · 2026-10-21
- Holon-S1 predicts gene activity and biomarkers from routine pathology slides — altryne · 2026-10-21
- Three reasons AI is not conscious and may never be, per Nature essay — anilkseth · 2026-10-21
- Aviation tracks every part's provenance — agents can reason over that graph — colinmcnamara · 2026-10-21
- SFT details: instruction/reasoning/code/math mix, synthetic + human traces, no epoching — TimDarcet · 2026-10-21
- Mid-training recipe disclosed: 32k→131k context extension, dropping logit distillation for tokens — TimDarcet · 2026-10-21