Main contents


Development of data fusion and diagnostics methods in weather station networks

One of the current trends in weather forecasting is nowcasting, i.e. doing short term and local forecasts with a time frame of a couple of hours and a spatial accuracy of a few kilometers. Nowcasting requires more dense measurement network of surface weather, along with remote sensing methods (Doppler radars, weather satellites etc.). For nowcasting research purposes, Finnish Meteorological Institute (FMI) and Vaisala have developed their own weather measurement network, Helsinki Testbed, which includes new Vaisala weather stations WXT510s.

FMI's Doppler radar in Korppoo, Finland (© Finnish Meteorological Institute)

DADA-project utilizes the Helsinki Testbed data to reach its two main objectives. The first one is data fusion of different weather sensors and local forecasting of convective phenomena. The second objective is maintenance strategy of an automatic surface weather station network.

The goals for the data fusion part are

  • Development of diagnostic algorithms of convective storms
  • Data fusion of measurements of different sources (e.g. surface stations, Doppler radars)
  • Short timeframe extrapolation methods for convective phenomena

Due to the variety and complexity of weather phenomena, DADA is concentrated on convective thunders, which can be distinctly found by different measurement sources. DADA will develop algorithms for estimation and prediction of life cycle of convective thunder cells.

The maintenance part of DADA is concentrating on

  • Fault identification in weather stations
  • Development of suitable fault diagnostic methods
  • Classification of errors, life span estimation and maintenance determination of weather stations
  • Development of maintenance strategy

Vaisala WXT510 weather station (© Vaisala Oyj)

Automatic sensor fault identification in weather stations has a difficult task - it must separate actual sensor faults from ordinary weather phenomena. The fault detection must be dependable. Since the weather measurement networks must be large scale, poorly designed fault detection and maintenance can be costly. DADA aims for smart maintenance strategies with life span estimation of weather stations.

DADA is a part of the Tillikka-project. Tillikka works on improving industrial competitiveness with condition monitoring.

The research is financed by Tekes (National Technology Agency of Finland). The research partner is Finnish Meteorological Institute and the industrial partner is Vaisala.


  • Heikki Koivo - person in charge
  • Vesa Hasu - project manager
  • Reino Virrankoski
  • Pekka Rossi
  • Markus Peura (FMI)