Weather Sources
Naola queries multiple weather services at the same time and combines their forecasts into a single consensus prediction. Here's how it works and which sources are used.
Current Sources
The free tier queries 3 sources. Naola Advanced adds 3 more for a tighter consensus, 6 in total.
| Provider | Forecast Model | Tier | Website |
|---|---|---|---|
| Open-Meteo | ECMWF (European model) | Free | open-meteo.com |
| OpenWeather | GFS (US model) | Free | openweathermap.org |
| Apple Weather | Blended (Apple) | Free | Apple Weather data sources |
| WeatherAPI.com | Blended | Advanced | weatherapi.com |
| Visual Crossing Weather | Blended | Advanced | visualcrossing.com |
| Blended (Google Maps Platform) | Advanced | developers.google.com/.../weather |
Home-screen widgets show the sources your plan includes, except Google, which appears only in the app.
How the Consensus Works
- All sources are queried at once: If one source is slow or down, the others still work. You always get a result.
- Forecasts are normalized: All temperatures are converted to Celsius and all wind speeds to m/s, so they can be compared fairly.
- A weighted median is computed: Sources powered by higher-quality forecast models (like the European ECMWF model) count more than blended models. This produces a more accurate consensus.
- Outliers are removed: If one source gives a wildly different reading, it's excluded from the final result to avoid skewing the consensus.
Confidence Colors
The colored badge on the weather card tells you how much the sources agree:
| Color | What It Means |
|---|---|
| Green | High confidence: sources agree within 1.5 degrees C |
| Yellow | Moderate confidence: some disagreement (within 3 degrees C) |
| Red | Low confidence: sources disagree by more than 3 degrees C |
Forecasts
The consensus approach also applies to forecasts:
- Daily forecast: Up to 7 days ahead, combining predictions from all sources
- Hourly forecast: Up to 48 hours ahead, for more granular planning
How Source Accuracy Is Measured
The Source Accuracy screen ranks weather sources country by country, by how close their forecasts came to what weather stations measured. It is a personal tool to help you choose a source. It is not official or scientific, and it does not certify any provider. None of the weather services named below is involved in this ranking or endorses it. Here is the full method, so anyone can repeat it:
- Stations: In each country, a fixed set of reference weather stations, run by that country's national weather service and listed below. They are the same for everyone: the ranking never depends on where you search, and Naola keeps no record of where you look for it.
- Forecasts: Every day at 08:00 UTC, Naola records each source's forecast high and low for the next day at each station, using the station's own calendar day, and the Naola Consensus of those forecasts. Apple Weather, Google and WeatherAPI.com are never recorded, because their terms do not allow keeping their forecasts, so they are never ranked. The Naola Consensus ranked here therefore combines only Open-Meteo, OpenWeather and Visual Crossing Weather, and can differ from the consensus on your weather screen. Forecasts for Germany, France, Austria, Czechia and Japan are recorded at 08:20 UTC instead.
- Measurements: Each station's measured maximum and minimum temperature for that day is read from its national weather service once published, usually a day later (two days for France, about three for the United States). A few services define the day differently: SMHI's runs from 18:00 to 18:00 UTC, GeoSphere Austria's from 18:10 to 18:00 UTC and ČHMÚ's from 20:00 to 20:00 UTC; MeteoSwiss's and DWD's run from midnight to midnight UTC, and JMA's is the calendar day in Japan. Météo-France takes the low from 18:00 to 18:00 UTC, ending on the day, and the high from 06:00 UTC on the day to 06:00 UTC the next day. ČHMÚ and Météo-France do not publish these two windows: Naola inferred them from the services' own hourly or 10-minute readings. Each window still holds the day's afternoon high and early-morning low. Services can still correct recent values for a few weeks; Naola keeps the first value it reads, and for Japan uses only values JMA marks as normal quality.
- Score: For each source, station and day, the error is the difference between the forecast and the measurement, for the high and for the low, ignoring the sign, in °C. A source's score for a period is the average of those errors over all its stations and days.
- Ranking: Within each country, sources are ranked by their average score over the chosen period (7, 14 or 30 days), lowest first. A country's ranking is shown once it has 14 days of scores from at least 5 stations (or from all of them, in a country with fewer). Countries added later, such as Germany, France, Austria, Czechia and Japan, appear on the same terms.
Countries and stations
- United States, 25 stations measured by NOAA: New York (Central Park), Boston, Washington (Reagan National), Atlanta, Miami, Detroit, Chicago (O'Hare), Minneapolis-St Paul, Nashville, New Orleans, Houston (Intercontinental), Dallas-Fort Worth, Denver, Albuquerque, Salt Lake City, Billings, Phoenix, Las Vegas, Los Angeles, San Francisco, Portland (Oregon), Seattle-Tacoma, Anchorage, Fairbanks, Honolulu.
- Sweden, 6 stations measured by SMHI: Stockholm (Observatoriekullen), Göteborg, Malmö, Umeå (airport), Luleå (Kallax airport), Kiruna (airport).
- Finland, 7 stations measured by the Finnish Meteorological Institute (FMI): Helsinki (Kaisaniemi), Turku (Artukainen), Tampere (Härmälä), Kuopio (Savilahti), Oulu (airport), Rovaniemi, Sodankylä.
- Denmark, 6 stations measured by DMI: Copenhagen (airport), Aalborg, Aarhus (airport), Odense (airport), Billund (airport), Bornholm (airport).
- Canada, 8 stations measured by Environment and Climate Change Canada (ECCC): Toronto (Pearson), Montréal (Trudeau), Vancouver (airport), Calgary (airport), Winnipeg (airport), Halifax (Stanfield), St. John's (airport), Yellowknife (airport).
- Switzerland, 6 stations measured by MeteoSwiss: Zürich (Fluntern), Geneva (Cointrin), Basel (Binningen), Bern (Zollikofen), Lugano, Sion.
- Germany, 8 stations measured by the German Weather Service (DWD): Berlin (Tempelhof), Hamburg (airport), Rostock (Warnemünde), Dresden (airport), Cologne/Bonn (airport), Frankfurt am Main (airport), Stuttgart (airport), Munich (city centre).
- France, 8 stations measured by Météo-France: Paris (Montsouris), Lille (airport), Strasbourg (airport), Brest (airport), Lyon (airport), Bordeaux (airport), Toulouse (airport), Marseille (airport).
- Austria, 7 stations measured by GeoSphere Austria: Vienna (Hohe Warte), Linz (city centre), Graz (airport), Klagenfurt (airport), Salzburg (airport), Innsbruck (airport), Bregenz.
- Czechia, 8 stations measured by the Czech Hydrometeorological Institute (ČHMÚ): Prague (airport), Cheb, Plzeň, České Budějovice, Liberec, Pardubice, Brno (airport), Ostrava (airport).
- Japan, 9 stations measured by the Japan Meteorological Agency (JMA): Sapporo, Sendai, Tokyo, Niigata, Nagoya, Osaka, Hiroshima, Fukuoka, Naha.
Station measurements: credits and licences
- NOAA: Menne et al. (2012), Global Historical Climatology Network - Daily (GHCN-Daily), Version 3, NOAA National Climatic Data Center, doi:10.7289/V5D21VHZ, accessed daily; United States stations only. US Government work, in the public domain.
- SMHI: Source: SMHI. Licence CC BY 4.0. Modified: daily maxima and minima used to compute forecast errors.
- FMI: Source: Finnish Meteorological Institute. Licence CC BY 4.0. Modified: daily maxima and minima used to compute forecast errors.
- DMI: Source: DMI (Danish Meteorological Institute). Licence CC BY 4.0. Modified: daily maxima and minima used to compute forecast errors.
- ECCC: Data Source: Environment and Climate Change Canada. Licence: ECCC Data Servers End-use Licence. Modified: daily maxima and minima used to compute forecast errors. Contains information licenced under the Data Server End-use Licence of Environment and Climate Change Canada.
- MeteoSwiss: Source: MeteoSwiss. Licence CC BY 4.0. Modified: daily maxima and minima used to compute forecast errors. MeteoSwiss does not endorse this ranking.
- DWD: Datenbasis: Deutscher Wetterdienst, Einzelwerte gemittelt. Licence CC BY 4.0. Modified: daily maxima and minima used to compute forecast errors.
- Météo-France: Source : Météo-France, Données climatologiques de base - quotidiennes. Licence: Licence Ouverte / Open Licence 2.0. Modified: daily maxima and minima used to compute forecast errors. Date of last update: shown with the ranking in the app.
- GeoSphere Austria: Datenquelle: GeoSphere Austria - https://data.hub.geosphere.at. Licence CC BY 4.0. Modified: daily maxima and minima used to compute forecast errors.
- ČHMÚ: Source: ČHMÚ (Czech Hydrometeorological Institute). Licence CC BY 4.0. Modified: daily maxima and minima used to compute forecast errors.
- JMA: 出典:気象庁ホームページ(https://www.data.jma.go.jp/stats/data/mdrr/index.html)。Naola が編集・加工(予報誤差の算出)。 Licence: Public Data License 1.0. Modified: daily maxima and minima used to compute forecast errors.
Attribution and Licences
- Open-Meteo: weather data by Open-Meteo.com, licensed under CC BY 4.0. Naola changes it: values are normalized and combined with other sources.
- OpenWeather: weather data and location search provided by OpenWeather, available under the Open Database License (ODbL).
- Apple Weather: weather data and alerts from Apple Weather, which draws on these data sources. Consensus values modify Apple's data by combining it with other sources. Apple Weather is a trademark of Apple Inc.
- WeatherAPI.com: weather data powered by WeatherAPI.com.
- Visual Crossing Weather: weather data provided by Visual Crossing.
- Google: Source: Includes weather data from Google. Naola includes Google Maps features and content, subject to the Google Maps End User Additional Terms of Service and the Google Privacy Policy.
- Copernicus (air quality, via Open-Meteo.com): Contains modified Copernicus Atmosphere Monitoring Service information 2026. Neither the European Commission nor ECMWF is responsible for any use that may be made of the Copernicus information or data it contains.
Copernicus citations
- Copernicus Atmosphere Monitoring Service (2020): CAMS European air quality forecasts. Copernicus Atmosphere Monitoring Service (CAMS) Atmosphere Data Store, DOI: 10.24381/a4005cee.
- Copernicus Atmosphere Monitoring Service (2021): CAMS global atmospheric composition forecasts. Copernicus Atmosphere Monitoring Service (CAMS) Atmosphere Data Store, DOI: 10.24381/04a0b097.
- METEO FRANCE, Institut national de l'environnement industriel et des risques (Ineris), Aarhus University, Norwegian Meteorological Institute (MET Norway), Jülich Institut für Energie- und Klimaforschung (IEK), Institute of Environmental Protection – National Research Institute (IEP-NRI), Koninklijk Nederlands Meteorologisch Instituut (KNMI), Nederlandse Organisatie voor toegepast-natuurwetenschappelijk onderzoek (TNO), Swedish Meteorological and Hydrological Institute (SMHI), Finnish Meteorological Institute (FMI), Italian National Agency for New Technologies, Energy and Sustainable Economic Development (ENEA) and Barcelona Supercomputing Center (BSC) (2022): CAMS European air quality forecasts, ENSEMBLE data. Copernicus Atmosphere Monitoring Service (CAMS) Atmosphere Data Store (ADS). URL: ads.atmosphere.copernicus.eu/datasets/cams-europe-air-quality-forecasts.