GribStream Blog
NBM OCONUS core grids now available: Alaska, Hawaii, Puerto Rico, Guam, and Oceanic
GribStream expanded NOAA's National Blend of Models (NBM) beyond CONUS with core grids for Alaska, Hawaii, Puerto Rico, Guam, plus the Oceanic domain.
On February 5, 2026, we expanded NOAA's National Blend of Models (NBM) coverage on GribStream beyond the CONUS core grid to include OCONUS domains (Outside Continental US) plus the Oceanic domain.
If you build anything that depends on high-resolution, calibrated NOAA guidance (disaster preparedness, infrastructure risk, energy, insurance, or operations), these regional grids often matter more than global "one size fits all" products.
What's included (coverage + resolution)
- CONUS core:
nbm(~2.5 km) (model page) - Alaska core:
nbmak(~3 km) (model page) - Hawaii core:
nbmhi(~2.5 km) (model page) - Puerto Rico core:
nbmpr(~1.25 km) (model page) - Guam core:
nbmgu(~2.5 km) (model page) - Oceanic core:
nbmoc(coarse, ~10 km near the tropics) (model page)
All six datasets are served from the AWS Open Data NBM archive and expose the same NBM core product structure (hourly runs with lead times extending out to 264 hours, depending on the field).
Why it matters
Many weather APIs are CONUS-centric. In practice, "NBM support" often stops at the CONUS core grid, and the Hawaii / Puerto Rico / Guam domains are exactly where you need reliable, local guidance when high-impact weather hits.
Adding OCONUS + Oceanic core grids enables:
- Natural disaster warning & prevention: tropical cyclones, extreme rainfall, high winds, and flood-risk monitoring for Hawaii, Puerto Rico, and Guam, plus severe winter and coastal hazards in Alaska.
- Operations in islands and territories: the same workflow you use for CONUS, but on the native grid for each domain.
- Marine/offshore analytics: wide-area conditions (wind/waves and related hazards) via the Oceanic core grid.
- Backtesting + ML training: consistent access to forecast history across regions without stitching together multiple providers.
Why it's a differentiator for GribStream
Competitors like Open-Meteo are great for general-purpose forecast APIs, but they typically don't expose these NBM regional core grids as first-class datasets with consistent query semantics and long forecast history.
GribStream's goal is the opposite: make the niche-but-important feeds easy to use, especially when your domain is public safety, risk, or critical infrastructure and you can't ignore OCONUS coverage.