WIND FORECASTS IN BARCARES

Generation of the last forecasts: 24 July 2017 05:05:48 (UTC+1)

ABOUT THIS PROJECT

An accurate local weather forecast

This model provides a local weather prediction of the wind on a windsurf spot in Le Barcarès (southern France). Three characteristics are generated: the mean speed, the gusts speed, the mean direction. This model is automatically updated up to 4 times every days and it produces hourly forecasts on a 7-days period available on several charts.


how to use it?

The gauges initially display the current prediction and they are automatically refreshed every 3 minutes. Use the timeline bar to get a forecast at any other time.


How does it work?

The outputs from the Global Forecast System (GFS) numerical model are compared on a long period with the actual wind reports provided by a station operated by Winds-Up. Predictive analytics techniques then enable to correct the flow of GFS data to provide a real-time improved local forecast. This approach is transposable to the prediction of any weather-dependant variable.


Target data source

provides a real-time access to a network of more than 120 anemometers and nearly 40 webcams installed on kitesurf and windsurf spots mainly in France, Spain and Italy. Winds-Up also offers weather forecast from several models on these spots and a unique database of weather records on several years for "big data" analysis.


Terms of use

All the data generated by this experimental model are freely available without warranty of any kind.


Consulter la version française

WIND FORECAST CHARTS

The 7-days forecast of the wind is compared with the GFS model raw outputs they are derived from. The following charts are all in UTC time.

FEEDBACK AND VALIDATION

These models have been created by exploiting actual weather records and it is therefore crucial to permanently validate them with the continuous flow of new reports. This approach will also enable to correct the predictions and improve the accuracy in the future.

The following charts give some examples of comparison between forecasted and actual values: