In 2019, ECMWF made decisive progress towards the operational use of wind data from the European Space Agency’s ground-breaking Aeolus satellite. ECMWF scientists worked with others to resolve data quality issues, improve data processing and correct biases. By the end of the year, they had shown that assimilating Aeolus data improves forecasts.

Aeolus© ESA/AOES Medialab

After more than two decades of development, ESA’s polar-orbiting Aeolus satellite was launched in August 2018. It carries just one large instrument – a Doppler wind lidar called ALADIN - which is the world’s first functioning lidar to provide profiles of horizontal line of sight winds from space. Observations from Aeolus help to fill a significant gap in the global observing system, as many parts of the globe, such as the tropics, oceans and upper troposphere, lack wind profile measurements.

ECMWF has been closely involved with the mission right from the design phase. In 2019, the Centre made major strides in testing Aeolus data within the Integrated Forecasting System (IFS) to:

  • understand data quality and biases and develop corrections,
  • prepare the data and analysis system for operational assimilation to improve estimates of the initial state of the Earth system at the start of forecasts,
  • investigate the effect of assimilating the data on medium-range weather forecasts.

ECMWF worked within the Aeolus DISC (Data Innovation and Science Cluster) closely with ESA, the German Aerospace Center (DLR), Météo-France, the Dutch national meteorological service (KNMI) and the software company DoRIT to identify and resolve any data quality issues and to contribute to improved data processing. The main problem from a data assimilation point of view was that the Aeolus data had biases that were larger and more complex to understand than was expected before launch.

Systematically comparing Aeolus winds with short-range ECMWF forecasts provided a very valuable method to assess the Aeolus wind biases. The method was also used to compute day-by-day bias corrections depending on the satellite’s position in orbit. Tests showed that applying such bias corrections improves the impact of the Aeolus data on forecasts.

Further work showed that biases in Aeolus wind data were closely correlated with tiny temperature variations across the 1.5 m diameter mirror which forms an important part of the Aeolus instrument. A 0.1°C change in temperature difference was found to lead to a change in wind bias of around 5 m/s.

These results enabled Aeolus engineers and scientists to start investigating why such temperature differences cause large wind biases and if the mirror temperatures can be controlled better. This led to the development of an improved bias correction method. A better understanding will be very valuable when designing and building a potential follow-on Aeolus instrument.

The impact of assimilating Aeolus data compared to other Earth system observations was also tested. The overall impact of Aeolus data on short-term forecasts in the southern hemisphere and the tropics was found to be comparable to that of data from some other major components of the global observing system, such as Atmospheric Motion Vectors, satellite-to-satellite radio occultation data, and satellite data from infrared sensors.

Crucially, inclusion of Aeolus data was found to bring clear improvements in medium-range weather forecasts. The improvements are most marked in the tropics and near the poles.

The positive results meant that by the end of 2019 ECMWF was testing the assimilation of Aeolus data in e-suite mode in preparation for operational implementation in January 2020.

Impact on forecasts
Impact on forecastsExample of the impact of assimilating Aeolus wind data on 2-, 3- and 5‑day forecasts for the period 2 August to 31 December 2019. The figure shows the relative change in root-mean-square (RMS) error of the vector wind forecast with the assimilation of Aeolus data compared to without it, verified against operational analyses. Hatching shows statistical significance at the 95% confidence level. Blue colours indicate forecast improvement due to Aeolus.

As early as 15 months after launch, ECMWF and several other numerical weather prediction centres have shown very large improvements in weather forecasts when Aeolus data is assimilated in test experiments. This is a success story of close collaboration between ESA, ECMWF, other weather prediction centres and all scientists involved.

Tommaso Parrinello, Aeolus Mission Manager at ESA