Aeolus validation and assimilation

The ESA Earth Explorer mission Aeolus is the first satellite worldwide carrying a Doppler Wind Lidar on-board, providing atmospheric wind profiles on a global basis up to heights of 30 km. Aeolus provides horizontal line-of-sight (HLOS) wind measurements. In the example of Fig.1 we show the coverage of Aeolus observations over 12 hours worth of orbits: the red colors over the North Atlantic ocean capture very strong jet stream winds in the upper troposphere.

The assimilation of Aeolus HLOS winds into numerical weather prediction models has improved forecasts quality across the board. Main examples are the European Centre for Medium-Range Weather Forecasts (ECMWF) and the German Weather Service (DWD), both of which collaborate with the “Numerical Weather Prediction and Data Assimilation” group (NWP and DA). NWP centres usually provide statistics on forecast performance, e.g. percentage improvements in forecasts assimilating Aeolus HLOS winds. These standard statistics summarize all the information globally, and don’t differentiate the different weather regimes. The efforts within the “NWP and DA” group are to unravel the atmospheric dynamics behind these improvements.

In the mid-latitudes, we focus on the jet stream and its undulations, that can organize into a wave packet (an organized succession of low and high pressure systems). These occasionally interact with transitioning tropical cyclones, which bring high amounts of energy and divergent circulation at the upper levels of the troposphere, and affecting the downstream evolution of the wave packet. These interactions are challenging to forecast with high precision. In the example in Fig.2, dashed black lines indicate forecast improvements by Aeolus assimilation into the operational ICON model from the DWD. Improvements follow a wave packet (orange/green shading) that interacted with a transitioning tropical cyclone (pink dots). In the aftermath of this interaction, the downstream development of the wave packet coincides with the location of forecast improvements. Aeolus thus provided valuable additional information in this unbalanced and highly energetic synoptic situation, that led to the improved forecast of the wave packet.

Precipitation is a difficult parameter to forecast and to evaluate due to its highly asymmetric distribution. In the example showed in Fig. 3, the forecast-skill-score method was used for precipitation stronger than 10 mm/day. Aeolus assimilation shows forecast error reductions of 5-10% for forecast done 7-10 days ahead of time, especially during the winter in each Hemisphere’s extratropics. It is thought this is due to the assimilated Aeolus HLOS constraining large-scale circulation e.g. the above-mentioned wave packets.

Ongoing work within the group also addresses the atmospheric dynamics behind forecast improvements in the tropics, looking at equatorial wave modes, the quasi-biennial oscillation in the lower stratosphere, and a focus on the tropopause region as well.