Validation of Aeolus Rayleigh and Mie winds using atmospheric radars in Arctic Sweden and in Antarctica

So far, validation of Aeolus winds for the polar regions has been based on the ECMWF global data assimilation and forecasting system (e.g. Rennie and Isaksen 2020).  Very few conventional upper air meteorological measurements (radiosondes, aircraft in-situ sensors) are available in the polar regions so the model’s accuracy is not well known in those regions.  There is a risk that different cloud conditions, surface reflectivities and summer daylight in these regions could lead to different performance of the space-borne lidar measurements. At the same time, accurate measurements over the polar regions would be a particular asset to global weather forecasting and climate monitoring as these regions are so poorly covered by other observations. We validate Aeolus Rayleigh and Mie winds by comparison with winds measured by two atmospheric radars, ESRAD and MARA, located at Esrange (68°N 21°E) in Arctic Sweden and at the Indian Antarctic station Maitri (71°S 12°E), respectively, for the period July - December 2019 when reprocessed data for baseline 10 with the telescope mirror temperature correction were available. Data were divided into two seasons: summer with 12 -24 hours direct sunlight and winter covering the rest of the time. Aeolus - radar collocation events are defined when the distance between Aeolus measurement swath and the radar sites is less than 100 km. We computed regression, bias, and standard deviation for the Aeolus winds in comparison with the radars. For Rayleigh winds the slope of regression line is not significantly different from 1, and bias is not significantly different from 0.  Random difference (std) is 4.4 m/s – 7 m/s. For Rayleigh winds at both locations and Mie winds at Esrange we did not find any statistically significant difference between ascending/descending orbits and seasons. However, at Maitri, Antarctica a few m/s bias is found for Mie winds in summer for ascending (evening) passes.