Data Sources
This page documents where WeatherRecall data comes from, how it is processed, and what quality controls are applied before insight text is shown. Transparent sourcing is essential for user trust and quality review.
Primary global source
WeatherRecall uses Open-Meteo historical archive endpoints for broad geographic coverage and consistent variable access. This provides core daily and hourly metrics such as temperature, precipitation, and wind, which are then transformed into user-facing comparisons and rankings.
Local station augmentation
Where station datasets are available, WeatherRecall can incorporate local observations to improve geographic realism. Station proximity matters: near stations better represent local microclimate behavior than distant ones. Blending logic prioritizes nearby signals while preserving consistency with baseline datasets.
Variable set and derived fields
- Direct variables: max/min temperature, precipitation, wind speed, pressure indicators, and cloud-related fields.
- Derived fields: anomaly deltas, percentile ranking, sunshine proxy, and weather-score context.
- Generated narrative text is constrained by available values and threshold checks.
Data processing pipeline
- Retrieve historical records for requested coordinates and time windows.
- Normalize fields into WeatherRecall internal structure.
- Compute baselines and historical distributions.
- Generate comparisons, records, and FAQ answers from measured values only.
- Store summary metrics for aggregate record databases and trend analysis.
Quality controls
- Missing values are surfaced as unavailable rather than estimated invisibly.
- Extreme claims require direct metric support.
- Rankings are derived from observed sample distributions.
- Contradictory metrics are not forced into narrative claims.
Update and interpretation guidance
WeatherRecall evolves as source quality and platform analytics improve. Users should treat historical outputs as evidence-based context, not deterministic future forecasts. For operational decisions, combine historical analysis with official current forecast services.
Related pages: Methodology, How It Works, FAQ.