How WeatherRecall works

1. Select a place and calendar date

The main search resolves a place to coordinates, then requests the same month and day across the selected previous years. Confirm the country and location before interpreting results: a familiar place name can refer to several towns. The map provides another way to choose coordinates. Browser geolocation, where offered, requires your permission.

2. Request the archive

WeatherRecall asks Open-Meteo for archived daily and hourly variables, with the provider resolving the local time zone. This is historical reanalysis or model-based archive data, not a live sensor at the clicked point. The main query does not explicitly select one fixed archive model. Nearby catalog station coordinates can also be queried through the same API; this does not import those stations' original observations.

3. Build summaries

Weather categories organise conditions, time-of-day panels summarise hourly values, and application-defined scores provide a compact comparison. The daily mean, maximum, total, and category are different measures. Rain at night may affect a daily total without describing the afternoon. A cloud-based sunshine proxy should not be reported as measured sunshine hours.

4. Interpret a frequency

Hypothetical example: eight qualifying days among ten retrieved years gives an 80% sample frequency. That is a description of those ten entries, not a calibrated 80% forecast for your future event. With a small sample, one additional wet year changes the percentage noticeably. Changing the lookback window can also change the result.

5. Broaden the question

Use location comparison for alternatives, monthly summaries for season screening, and annual reports to investigate a particular year. These tools do not share every aggregation rule, so verify dates, variables, and baselines before comparing their numbers. Global ranking tables are a sample of recorded lookup activity, not world-record certification.

6. Verify consequential claims

Keep a note of coordinates, units, time zone, date window, and retrieval date. Consult Open-Meteo documentation and ECMWF reanalysis guidance; use NOAA station archives for independent comparisons when suitable. Source disagreements deserve investigation rather than invented replacement observations.

This project publishes educational synthesis, not certified forecasts or personalized safety advice. Official current warnings, venue constraints, and qualified local advice take precedence. See methodology, data sources, and frequently asked questions.