Met Office Blended Probabilistic Forecast – Global Gridded Percentiles

Met Office Blended Probabilistic Forecast data provides global weather forecasts as percentiles at approximately 20km grid resolution. The data is derived from operational Numerical Weather Prediction ensembles and nowcasts, blended and calibrated using the IMPROVER pipeline. This is one of eight such probabilistic forecast products published by the Met Office on AWS.

Use Cases

• Analyzing the range of possible weather outcomes based on ensemble percentiles.
• Verifying forecast reliability and spread–skill using calibrated probabilistic data.
• Studying global weather patterns at a 20km resolution grid.
• Investigating precipitation forecast uncertainty with expert interpretation of percentiles.

Strengths

• Global spatial coverage at approximately 20km resolution.
• Data is calibrated and verified using spread–skill and reliability checks via the IMPROVER pipeline.
• One of eight related probabilistic forecast products, indicating a systematic output.

Limitations

• Precipitation percentiles are noted as potentially misleading for non-experts due to spatial noise and interpretation challenges.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and specific temporal coverage are unknown, which may limit suitability assessment.

Provenance

Source

Met Office

Collection Method

Derived from operational NWP ensembles and nowcasts, blended and calibrated using the IMPROVER system.

Time Range

null

Freshness

Information is correct as of April 2026, but parameters may change.

Geography

Global