GribStream users and weather data fans: big news.
GribStream just added full support for NOAA's GEFS ensemble model. This means you can now access all individual GEFS ensemble members, not just the ensemble mean.
What’s new:
- You get every perturbed forecast member (30 members +1 control)
- Available globally at 0.25° grid, 4 runs per day (00, 06, 12, 18 UTC), out to 10 days
- Access via the same simple API, just set model=gefsatmos and specify the members array if you want specific members
Why this matters:
- Allows you to do true probabilistic and uncertainty analysis
- Calculate spread, percentiles, standard deviation, member agreement, etc
- Build weather risk tools, insurance models, energy applications, or advanced meteorological research on a budget
How to use:
- See the [GribStream GEFS info page](https://gribstream.com/models/gefsatmos) for details, variable list, and example queries
- All endpoints now support model=gefsatmos and an optional members parameter (example: "members": [2,3,4])
- Full documentation and OpenAPI spec have been updated
Tips:
- Data volumes increase if you pull all members — filter by variables, region, or members as needed
- Works worldwide, not just CONUS
- Ensemble members are indexed starting from 0 (control)
This is a big step for anyone who needs more than just the mean forecast. If you use ensembles for weather analytics, power grids, insurance, ag-tech, or just want to play with real ensemble weather data: check it out.