Weather (also called meteorological) measurements have created a scientific base for determining and comparing weather information, and have been developed rapidly since the 18th century. Measurements can be made for meteorological variables, such as temperature, windspeed, precipitation, and others.

The measurements come mainly from:

  • weather stations, as displayed for example on the temperature °C maps;
  • weather balloons ("soundings");
  • buoys (on the water).

And for more specific data (cloud cover, water vapour, etc.):

While until the late 20th century, most measurements needed to be conducted manually, most of them have been automated by 2010, enabling measurements of more variables, in more frequent time intervals, with more precision and with easier access to the data.

Nevertheless, measurements remain available only for less than 1% of the Earth's surface. The graph shows the coverage of the land surface by measurements, assuming a measurement is valid for an area with a diameter of 50, 25, 12, 3 or 1 kilometre.

Land coverage with weather stations

Land coverage with weather stations

In case only weather stations are used, the weather information coverage of the Earth's land surface is limited to less than 50%, even if a 50 km diameter (25 km radius around each weather station) is assumed. If weather conditions (such as temperature, wind, precipitation, etc.) change every 3 km, then less than 1% of the land surface is covered by measurements. In Europe, only 0.3% of the land are covered by weather stations with a surrounding range (grid) of 3 km, and only 0.04% in Africa. If a global grid of 50 km is used, then 60% of Europe and only 8% of Africa are covered. 

With modeling, 100% of the world can be covered, and many events occurring in every place can be simulated, thanks to NMM technology. Precision of the information generated depends on the resolution and the variables calculated. Through the use of weather models, we can now understand weather conditions in most places of the Earth much better than it was possible in the past. Through the daily calculation of models, it is possible to visualise the weather situation in any place in the world.

Measurements are used to initialise models, to validate the model output at places where reliable measurements are available and to improve specific model output, if a better accuracy for one determined point is required. meteoblue continuously compares measurement data with the results of the model, and applies post-processing to improve some variables. If the difference between the measurement and the forecast is larger than expected, please inform us.