About Station Climate
Station Climate is a simple way to see what the weather has really been doing at one place over the long run. Pick a station, anywhere in the world, and get a clear picture: the hottest and coldest days ever measured there, how the yearly temperature has drifted over the decades, how precipitation has changed and how many tropical or frost days a typical year brings now compared with the past.
Why we built it
Climate change is usually described in global averages. Those numbers are correct but abstract; nobody lives in a global average. Weather stations have been writing down the temperature every day for a century or more, and that record is the most honest answer to the question "has it really changed here?". We wanted that answer to be one search away, for 48,951 stations, without charts you need a degree to read.
What you will find on a station page
- Records: the hottest and coldest days in the whole measurement history, with the full list of the top 100.
- Warming stripes: one stripe per year, blue for cooler, red for warmer than the 1961–1990 average.
- Temperature change: yearly departures from the long-term average with an 11-year running mean and the trend per decade.
- Precipitation: yearly totals compared with the long-term average.
- Climate normals: the 30-year monthly averages that meteorologists use as reference periods.
- Climate indices: summer, tropical, frost and ice days per year, and the number of rainy days.
Where the data comes from
Everything on this site is built from published observations of national weather services, combined from open datasets: Meteostat (CC BY 4.0), NOAA GHCN-Daily and GHCN-Monthly (public domain), SHMÚ open data (CC BY 4.0) for Slovakia and ČHMÚ open data (CC BY 4.0) for the Czech Republic. When several sources report the same day, national services take precedence. Values from forecast models are never used for records, and obvious measurement errors are filtered out: a day is dropped when it lies far outside the station's own climatology (or that of its nearest neighbour for short series), and maximum temperatures above 54.4 °C, the highest reliably measured air temperature, are treated as errors. Disputed historical readings such as the 56.7 °C at Death Valley in 1913 therefore do not appear here. Extreme values are also compared with neighbouring stations on the same day: a reading several degrees beyond anything measured within 100 km is treated as a sensor fault rather than a record. Where a station reports only daily minimum and maximum, the daily mean is estimated as their average and marked as derived.
Datasets often list the same site several times (an airport, its synoptic station and a climate station a few hundred metres apart). Stations within about 2 km of each other are merged into one page under the station with the longest record, and rain gauges without a thermometer are attached to the nearest temperature station within 5 km so that its page also shows precipitation. Gauges with no such neighbour have their own page with precipitation only.
Measurements are not homogenised: a station may have moved or changed instruments during its history, and such changes can show up in the long-term trend. Treat the trends as what the station recorded, not as a verdict on the regional climate.
What to expect from the data
- Not every station has everything. Many stations report only daily maximum and minimum, some only monthly means, others only precipitation. Charts and indices appear only where the underlying data exists, and a "days measured" count tells you how complete a record is.
- Some stations lag behind. National services publish quality-checked observations with a delay, which for some stations can be several months or even half a year. The data period shown on each station page is the honest range of what we have; a station that ends a few months ago is usually just waiting for its next data release, not closed.
- Gaps are normal. Wars, relocations and budget cuts leave holes in even the best records. Yearly values are computed only for years with enough months, and yearly indices only for years with at least 300 measured days, so a gap shows up as a missing year rather than a distorted one.
Fair use
Station Climate is free to use. If you reuse a chart or a number, please link back to the station page and credit the original data providers named in the footer.
Contact
Found a wrong value, a missing station or a better data source? Write to info@stationclimate.com.