Rayleigh wind retrieval for the ALADIN airborne demonstrator of the Aeolus mission using simulated response calibration

摘要

Aeolus, launched on 22 August in 2018, is the first ever satellite to directly observe wind information from the surface up to 30 km on a global scale. An airborne prototype instrument called ALADIN airborne demonstrator (A2D) was developed at the German Aerospace Center (DLR) for validating the Aeolus measurement principle based on realistic atmospheric signals. To obtain accurate wind retrievals, the A2D uses a measured Rayleigh response calibration (MRRC) to calibrate its Rayleigh channel signals. However, differences exist between the respective atmospheric temperature profiles that are present during the conduction of the MRRC and the actual wind measurements. These differences are an important source of wind bias since the atmospheric temperature has a direct effect on the instrument response calibration. Furthermore, some experimental limitations and requirements need to be considered carefully to achieve a reliable MRRC. The atmospheric and instrumental variability thus currently limit the reliability and repeatability of a MRRC. In this paper, a procedure for a simulated Rayleigh response calibration (SRRC) is developed and presented in order to resolve these limitations of the A2D MRRC. At first the transmission functions of the A2D Rayleigh channel double-edge Fabry-Perot interferometers (FPIs) in the internal reference path and the atmospheric path are characterized and optimized based on measurements performed during different airborne and ground-based campaigns. The optimized FPI transmission functions are then combined with the laser reference spectrum and the temperature-dependent molecular Rayleigh backscatter spectrum to derive an accurate A2D SRRC which can finally be implemented into the wind retrieval. Using dropsonde data as a reference, a statistical analysis based on a dataset from a flight campaign in 2016 reveals a bias and a standard deviation of line-of-sight (LOS) wind speeds derived from a SRRC of only 0.05 and 2.52 m s(-1), respectively. Compared to the result derived from a MRRC with a bias of 0.23 m s(-1) and a standard deviation of 2.20 m s(-1), the accuracy improved and the precision is considered to be at the same level. Furthermore, it is shown that the SRRC allows for the simulation of receiver responses over the whole altitude range from the aircraft down to sea level, thus overcoming limitations due to high ground elevation during the acquisition of an airborne instrument response calibration.