How Weather Stations Work

By WeatherRecall editorial project. Date reviewed: 2026-10-06. Last updated: 2026-10-06.

A weather station is a coordinated measurement site, not simply a thermometer connected to the internet. Its instruments convert physical changes into readings, a logger assigns times, and an operator maintains the equipment and surroundings. The resulting series describes conditions at that exposure. Knowing the station's design helps determine whether its record is suitable for a garden, a city, an airport, or a regional comparison.

What the instruments actually sense

An electronic thermometer commonly uses a material whose electrical resistance varies with temperature. Its enclosure should shield it from solar heating while allowing air movement. A thermometer exposed on a sunlit wall can accurately report its own heated surroundings while failing to represent standard outdoor air temperature. Shelter design and ventilation therefore matter alongside sensor specifications.

Humidity sensors respond to moisture, while pressure sensors respond to atmospheric force. Wind instruments may use rotating cups and a vane, or acoustic measurements without moving parts. A rain gauge collects water; tipping-bucket designs count small measured increments, while weighing gauges track accumulated mass. Different mechanisms introduce different failure modes, including blocked funnels, sticking bearings, and sensor contamination.

Measurement height belongs in the variable definition. The Open-Meteo historical documentation, for example, distinguishes air temperature at two metres from wind at ten metres. Those are definitions of its model variables, not proof that a physical station exists at the requested point, but they demonstrate why comparing undefined heights is unsafe.

From a sensor signal to a daily value

A logger samples instrument signals, converts them using calibration information, and stores values against a clock. It may retain short-interval readings and transmit longer averages. Losing a communications connection does not necessarily mean observations were lost: a functioning local logger may later backfill them. Losing power to the logger or sensor is a different problem.

Sampling, averaging, and reporting are separate intervals. A gust is a short-duration peak, not the same quantity as a sustained wind average. A daily high can be calculated from frequent samples or from hourly reports; the latter may miss a transient extreme. Metadata should describe these intervals before station records are compared.

NOAA GHCN-Daily distributes daily station summaries and links inventories containing identifiers, coordinates, elevations, and element coverage. A daily archive does not by itself reveal every detail of the original logger. For equipment-specific questions, consult the station operator's historical metadata.

Illustrative example: interpreting bucket counts

Illustrative example only; all counts and rainfall amounts below are invented, not real observed data. Assume a hypothetical tipping bucket records 0.2 mm per tip. During a complete six-hour interval it records 37 tips. The accumulated depth is 37 × 0.2 = 7.4 mm. The interval-average intensity is 7.4 ÷ 6 = 1.23 mm per hour.

That average says nothing about the wettest minute. If all the water arrived during half an hour, its average intensity during that burst would be 7.4 ÷ 0.5 = 14.8 mm per hour. Without tip timestamps, the burst cannot be reconstructed. Reporting the six-hour average as the storm's maximum intensity would change the meaning of the measurement.

If the funnel was blocked for part of the interval, neither arithmetic nor extra decimal places restores the missed rain. Nor should a partial bucket left at the interval boundary automatically count as an additional complete tip. Instrument resolution and interval boundaries constrain the result.

Why maintenance belongs in the record

Routine inspection checks levelling, drainage, shielding, power supply, clock settings, and encroaching vegetation. Calibration compares an instrument against a reference under controlled conditions. Field checks look for problems that a laboratory test may not reproduce, such as sunlight on a damaged screen or turbulence around a newly erected building.

Replacing a sensor can introduce an offset even when both units meet their specifications. Moving a station changes exposure. A maintenance log lets analysts separate these changes from weather variability. For long-term climate work, WMO's climate overview emphasizes quality-controlled, comparable observations and international observing standards. Continuity requires more than a long list of dates.

Choose by suitability, not distance alone

Illustrative station-selection example; these sites are fictional. Station A is 3 km from a coastal town but sits 400 metres above it. Station B is 12 km away at a similar elevation beside the same coast. B may better represent typical temperature, while neither necessarily captures a shower over the town. A distance ranking alone cannot settle the choice.

Match terrain, elevation, coastal exposure, and urban setting.
Check the required variable's actual coverage, not just the station's opening date.
Look for relocations, observation-time changes, and instrument replacements.
Retain the station identifier so the selection can be repeated.

Use the city tool to establish location context and the comparison tool to explore differences. Check methodology before assuming a displayed value comes from the nearest station.

What a station cannot promise

A reliable station measures one exposure, not every street around it. Urban surfaces can stay warm at night while an open field cools. Wind behind a building differs from wind over open ground. Keep these representativeness limits separate from instrument faults: a genuine local difference is not automatically bad data.

For historical reporting, describe a station value as evidence from that site and interval. Do not infer exact conditions at an unmonitored address, and do not use an old observation as a current safety forecast.

References

About the author

WeatherRecall editorial project prepares educational summaries using the listed public documentation. It is not a credentialed meteorologist or a station operator reporting original field research. The hypothetical gauge and site comparisons explain measurement concepts rather than documenting actual stations.