Validation of the Aeolus L2A products with the eVe reference lidar measurements from the ASKOS/JATAC campaign

Abstract

Aeolus was an ESA Earth Explorer satellite mission launched in 2018 with a lifetime of almost five years. The mission carried the Atmospheric Laser Doppler Instrument (ALADIN), a Doppler wind lidar for providing wind profiles on global scale and also vertically resolved optical properties of particles (aerosols and clouds) using the high spectral resolution lidar technique. To validate the particles' optical properties obtained from Aeolus as Level 2A products, the eVe lidar, ESA's reference system for the calibration and validation of Aeolus mission, has been deployed at the ASKOS campaign in the framework of the Joint Aeolus Tropical Atlantic Campaign (JATAC). ASKOS is the ground-based component of JATAC where ground-based remote sensing and in-situ instrumentation for aerosols, clouds, winds and radiation observations has been deployed at Cado Verde during summer 2021 and 2022 for the validation of the Aeolus products. The eVe lidar is a combined linear/circular polarization and Raman lidar specifically designed to mimic the operation of Aeolus and provide ground-based reference measurements of the optical properties for aerosols and thin clouds. The eVe lidar measurements can be used for the validation of the Aeolus L2A products while taking into consideration the ALADIN's limitation of misdetection of the cross-polar component of the backscattered signal. As such, in this validation study the cloud-free Aeolus L2A profiles obtained from the Standard Correct Algorithm (SCA), the Maximum Likelihood Estimation (MLE), and the AEL–PRO algorithms of Baseline 16 are compared against the corresponding cloud-free Aeolus-like profiles from eVe lidar, which are calculated using the retrieved particle circular depolarization ratio profile from eVe in order to account for the lack of the detection of the circular cross-polar backscatter component from Aeolus. The analysis focuses on the 14 collocated measurements between eVe and Aeolus during the nearest Aeolus overpass from the ASKOS site passing within a radius of 100 km from the site and within a temporal window of ± 1.5 h from the overpass time. The validation results reveal good performance for the co-polar particle backscatter coefficient, being the most accurate L2A product from Aeolus with absolute systematic and random errors up to 0.37 and 2 Mm⁻¹ sr⁻¹, respectively. The particle extinction coefficient also performs good with absolute systematic and random errors up to 53 and 183 Mm⁻¹, respectively. The co-polar lidar ratio seems to be the noisiest L2A product with extreme error values and variability. The observed discrepancies between eVe and Aeolus L2A profiles increase at lower altitudes where higher atmospheric loads (molecules and aerosols) are encountered leading to increased noise levels in the Aeolus retrievals due to enhanced laser beam attenuation, and greater atmospheric variability (e.g. Planetary Boundary Layer inhomogeneities) are typically encountered. Overall, this study underlines the strengths of the optimal estimation algorithms (MLE and AEL–PRO) with consistent performance and reduced uncertainties, while the standard inversion algorithm (SCA), which was originally developed, could be further improved particularly in the retrieval of the particle extinction coefficient and lidar ratio. In addition, the SCAmid-bin resolution profiles outperform the corresponding SCAnormal-bin as expected, since mid-bin resolution is obtained when averaging the values from two consecutive SCAnormal height bins.