The Weather File Feed, providing all the same weather parameters as the Weather API, is available as XML and delimited ASCII.
Vaisala provides raw weather data in an easy to use format for customers who want to design their own graphics, or who just need raw weather data for other purposes. Readymade Global or National packages are available - both with two different levels of service: Standard and Premium. For special purposes, custom feeds are also available upon request. Depending on the package chosen the feeds contain weather forecasts for up to 14 days, and current conditions data from relevant local weather stations.
The data is updated on a continual basis. With a Vaisala Weather File Feed your weather data will always be up to date.
Pricing is based on the package chosen and the end usage of the data. Any customer requested tailoring is charged separately. Graphical symbol sets are available for visualizing the weather data.
A Weather File Feed can contain weather information for more than 80.000 locations worldwide. Please note that the actual number of locations and pricing is subject to negotiations.
The Weather Feed is based on customer using HTTP polling to monitor forecast and observation files, which update at different intervals. Location metadata, such as coordinates and names in various languages, is provided in a separate static file. The files are either XML or delimited ASCII raw data, which is language independent. We recommend that larger batch files be gzipped delimited ASCII, since typically the delimited ASCII files are faster and easier to parse and gzip compression greatly reduces transfer times. For convenience (intended for smaller files) we can also provide the data as uncompressed XML or RSS. Vaisala reserves the option to limit traffic to only predefined IP addresses.
The exact update time may vary slightly from day to day, and there can be extra updates. To optimize data freshness and to create a robust system, please adhere to the following polling method:
The HEAD of a file is typically only hundreds of bytes and creates negligible traffic. For reference regarding the HEAD request, see http://www.w3.org/Protocols/rfc2616/rfc2616-sec9.html section 9.4, HEAD. Vaisala does not endorse any particular tools, but simply running wget in mirroring mode accomplishes the above.
The Vaisala Weather File Feed provides forecasts for up to 80.000 named locations. The needed locations can be chosen in your feed from among global and country packages. The global package contains 20.000 locations from all over the world. For dense local locations, the country packages offer more locations within the countries. The number of locations depends on the country. Some examples of the amounts are listed below.
Some examples of country location sets:
• Russia 3.900
• Finland 400
• Sweden 300
• Germany 4.600
• France 5.700
• United Kingdom 2.300
• United States 9.800
• Spain 2.900
• Italy 5.000
Global and National packages can be combined. For instance, Standard Global package with 20.000 locations including 352 UK locations UK Standard package with 2300 UK locations (all the locations included in the Global package and over 1800 additional locations). This combination provides complete coverage of the UK and all major locations globally.
Depending on the package chosen (Standard, Premium), the feed contains current conditions data from relevant local weather stations and weather forecasts for up to 14 days.
Current weather conditions are shown for the named locations based on the nearest weather stations. Vaisala gathers the observation data only from the official SYNOP and METAR weather stations to ensure high-quality data. The file is updated at least every 15 minutes, but the update frequency of any given observation station's data depends on the transmitting frequency of the station, which typically varies between 10 minutes and 3 hours; there are also some manual stations that do not measure during the night. When we choose which stations to use, we try to ensure that the weather conditions are as close as possible to the ones at the location they represent. Optionally, the feed provides also observations from the past day for selected stations and parameters.
With hourly forecast data, you are able to provide weather forecasts in one-hour-steps. Hourly forecast is adequate to show a detailed weather forecast of the current day as a table or graphical module. Forecasts are updated at least four times per day.
Daily Overview Forecast
The daily forecast is a 24-hour summary of the day. The data is provided for the next 14 days, including the current day. Forecasts are updated at least four times per day.
For the next five days, the feed contains all the most consumed weather parameters in 3- or 6-hour time steps depending on the chosen service level.
The following parameters are optionally provided for the next 30 days.
The service is delivered in up to four separate files, which are detailed below.
Vaisala provides graphical icons for visualizing the weather codes, which are (C) Vaisala but can be freely used to visualize the weather data.
This file contains summary-level weather parameters that are applicable for the entire day, such as the day's maximum temperature. Forecasts are updated at least four times per day.
This file contains snapshot-like weather parameters for the times specified, such as temperature, wind speed, etc. This data could, for instance, be used to plot a graph of a weather parameter as a function of time. Forecasts are updated at least four times per day.Current conditions
For each location, this file contains data from the closest relevant weather station at the time, if available. The file is updated at least every 15 minutes, but the update frequency of any given observation station's data depends on the transmitting frequency of the station, which typically varies between 30 minutes and 3 hours, but there are also some manual stations which do not take measurements during the night. When we choose which stations to use we try to ensure that the weather conditions are as close as possible to the ones at the location they are used to represent. To do this, we weigh the following factors:
Due to the above factors, a more close-by or recent observation might not be used, if it only contains temperature. We might opt for a slightly older observation from another station if it contains more weather parameters. This process is dynamic, so it is possible for the weather station chosen for a given location to change, as new data becomes available. One example situation where a station can change frequently happens when a close-by station transmits rich data during day time but ceases transmitting during the night, which is typical for smaller airports. In such a case a different station would be selected for night time.
Please note that in some parts of the world the observation network is really sparse, such as in Africa, parts of the Middle East, and South America, and especially during night time. The closest weather station with fresh data can at times be over 500 km away. In such cases, it is possible that the current conditions will be completely missing from the feed.
While the forecast data will always be there, please be prepared to handle missing current conditions (observations) information, since missing instrumentation or instrument malfunctions may affect data. Any of the parameter values can be missing/empty, and even the entire obs tag (in XML) can be missing. In delimited ASCII empty data is denoted by an empty string between the delimiters. In the case of missing data, it is recommended to simply adapt layout so that white space is shown instead. Texts like "No data" are NOT RECOMMENDED, as they are often interpreted by end-users as a service malfunction, which is not the case.