NOAA RRFS: Prototype High-Resolution Weather Forecasts for North America

A prototype of NOAA's next-generation Rapid Refresh Forecast System (RRFS), scheduled for operational implementation in 2026. The system provides hourly deterministic forecasts on a 3 km grid covering North America, Alaska, Hawaii, and Puerto Rico, out to 84 hours. It is underpinned by the Unified Forecast System (UFS) and developed collaboratively across NOAA, academia, and research institutions.

Use Cases

• Developing and testing initialization methods for RRFSv2 based on the prototype gridded data.
• Analyzing severe convective weather patterns based on the 3 km resolution, rapidly-updated forecasts.
• Forecasting renewable energy output based on high-frequency weather guidance for specific regions.
• Studying heavy precipitation and winter weather events using the ensemble prediction outputs.

Strengths

• Provides forecasts on a high-resolution 3 km grid covering multiple geographic domains.
• Offers hourly deterministic forecasts out to 84 hours at four main cycles per day.
• Data is provided in the standard GRIB2 format for meteorological data.

Limitations

• The prototype output is not operational, is not under 24x7 monitoring, and may be delayed or missing.
• Outputs will change, and the dataset description states significant changes will be updated.
• Column-level documentation is absent; field semantics must be inferred from file naming conventions.

Provenance

Source

National Oceanic and Atmospheric Administration (NOAA).

Collection Method

Output from the prototype Rapid Refresh Forecast System (RRFS) model.

Time Range

Real-time and retrospective parallel output; cycles include a prototype date of 14 January 2026.

Freshness

Real-time prototype data; update frequency is tied to forecast cycles (00, 06, 12, 18 UTC).

Geography

North America, with subsets for the contiguous United States (CONUS), Alaska, Hawaii, and Puerto Rico.