Latest news: 2016

How to make weather and energy forecasts better?

20.11.2016 12:46

Wind power forecasts produced by FMI during "Rauli" storm show promising agreement with the actual production figures.

Photo: Eija Vallinheimo

Photo: Eija Vallinheimo

A historical event occurred during 26-28 August 2016 when the storm "Rauli" hit Finland and the total generated wind power fixed by Fingrid exceeded 1000 MW for the first time. The total peak power was close to 1200 MW.

Over the last decade, renewable energy sector of Finland was actively growing. The increasing contribution of weather-dependent energy sources in energy market makes the role of weather forecasts even more essential. In particular, new energy market tools that are being developed within BC-DC, need power predictions as input information.

Developing reliable weather forecasts to these tools is the key objective of FMI in the project. To reach this goal FMI focuses on developing the areas that include three necessary steps in numerical weather forecasting:

  1. Analysis of weather observations

Every weather forecast requires the information on current weather conditions. This information comes routinely from synoptic weather stations, balloon measurements, commercial airplanes and from meteorological satellites and radars among others. The quality of weather forecasts thus depends to large extent on the spatial coverage of weather observations. However, the distances between meteorological stations are large and the data from the radars and satellites are not used nowadays in optimal way. To improve the accuracy of its weather forecasts FMI focuses on developing the methods to better use the information especially from radars and satellites.

     2. Numerical Weather Prediction

Weather forecasts with the range from few hours to few days are based on the results of Numerical Weather Prediction (NWP) systems. These systems processes observed data and by simulating dynamical and physical processes in the atmosphere and underlying surface produce the forecast of meteorological and other physical quantities. The computations are done on numerical grid that splits the atmosphere into a multitude of grid cells – similar to computer graphics.

Presently, FMI runs two NWP systems on its Cray-XC30 supercomputer– the one for large domain and coarse resolution, and the other one for smaller sub-domain around Finland and finer resolution. For better weather forecasting FMI works on improving description of physical processes in NWP systems. Spatial resolution of NWP systems running at FMI has continuously increased making weather forecasts more detailed and there are no plans to cease this process.

FMI also develops ensemble prediction system, which in addition to one predicted value will produce an ensemble of other possible values of a meteorological parameter. From energy production point of view it will give the range of uncertainty  in power values of the day that is very important particularly for wind energy operators.

  3. Postprocessing

The results of NWP systems are post-processed into the weather maps and variety of different products. To produce wind and solar power forecasts, FMI has been developing conversion tools that compute electric power from predicted wind speed and solar radiation. Currently, FMI can compute wind power for 70 wind farms in Finland, for which there is available information on the model of wind turbines and their specifications.  

During "Rauli" storm, FMI predicted around 81 % on average of the total power as compared to Fingrid's data. One possible reason for this underestimation is limited number of wind farms for which FMI computed the power: The total power capacity of the wind turbines included in FMI's calculations is just below 1000 MW while Fingrid estimates a total wind power capacity installed in Finland of around 1300 MW. The other possible reasons are lower predicted wind speed and uncertainties in production data. This is one of the subjects of FMI for the study.

"Rauli" storm case: 26-28 August, during which Fingrid fixed over 1000 MW of the total generated wind power in Finland for the first time.


Text by: Evgent Atlaskin

About the writer:

Evgeny Atlaskin works at the Finnish Meteorological Institute and has a PhD in meteorology. He graduated with master's degree from the faculty of meteorology of the State Hydrometerological Universtity in St. Petersburg, Russia in 2004, and continued his studies as a PhD student. At FMI he works mainly in the field of Numerical Weather Prediction for wind energy and road weather applications. 

Evgeny Atlaskin together with other FMI's researchers contribute to BCDC Energy research project as BCDC Weather Team:

See also Energy weather forecast service online: