EGU General Assembly 2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

An Assessment of the Impact of Aeolus Doppler Wind Lidar Observations for Use in Numerical Weather Prediction at ECMWF

Michael P. Rennie and Lars Isaksen
Michael P. Rennie and Lars Isaksen
  • European Centre For Medium-Range Weather Forecasts, Reading, United Kingdom (

The European Space Agency’s Aeolus mission, which was launched in August 2018, provides profiles of horizontal line-of-sight (HLOS) wind observations from a polar orbiting satellite.  The European Centre For Medium-Range Weather Forecasts (ECMWF) began the operational assimilation of Aeolus Level-2B winds on 9 January 2020 in their global NWP (Numerical Weather Prediction) model, 1 year and 4 months after the first Level-2B wind products were produced in near real time via ESA’s ground processing segment.  This achievement was possible because of the production of good data quality, which was met through a close collaboration of all the parties involved within the Aeolus Data Innovation and Science Cluster (DISC) and via the great efforts of ESA, industry and ground processing algorithms pre- and post-launch.
Through the careful assessment of the statistics of differences of the Aeolus winds relative to the ECMWF model, the Level-2B Rayleigh winds were found to have large systematic errors.  The systematic errors were found to be highly correlated with ALADIN’s (Atmospheric Laser Doppler Instrument) primary mirror temperatures, which vary in a complex manner due to the variation in Earthshine and thermal control of the mirror.  The correction of this source of bias in the ground processing is underway, therefore in the meantime a bias correction scheme using the ECMWF model as a reference was developed for successful data assimilation; the scheme will be described.  
We will present the results of the Aeolus NWP impact assessment which led to the decision to go operational.  Aeolus’ second laser (FM-B, available since late June 2019) provides statistically significant positive impact of moderate to large amplitude, of similar magnitude to some other important and well-established observing systems (such as IR radiances, GNNS radio occultation and Atmospheric Motion Vectors).  Observing System Experiments demonstrate reduction of forecast errors in geopotential and vector wind of around 2% in the tropics and 2-3% in the southern hemisphere for short-range and medium range forecasts (up to day 10).  This positive impact is particularly impressive given that Aeolus provides less than 1% of the total number of observations assimilated, showing the value of direct wind observations for global NWP.

How to cite: Rennie, M. P. and Isaksen, L.: An Assessment of the Impact of Aeolus Doppler Wind Lidar Observations for Use in Numerical Weather Prediction at ECMWF, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5340,, 2020


Display file