Validation of EarthCARE/ATLID aerosol profiling products with ground-based PollyNET lidars – case studies

Abstract. The satellite mission EarthCARE (Earth Cloud, Aerosol, and Radiation Explorer) of the European Space Agency (ESA) and the Japan Aerospace Exploration Agency (JAXA) was successfully launched in May 2024. The satellite has four instruments on board, namely a high-spectral-resolution lidar called ATLID, a Cloud Profiling Doppler Radar (CPR), a Multi-Spectral Imager (MSI), and a Broad-Band Radiometer (BBR). ATLID provides for the first time directly measured profiles of the extinction and backscatter coefficient (and thus lidar ratio) together with the depolarization ratio at 355 nm from space. Since the start of the measurements, several updates in the ESA’s processing chain have been made resulting in different base-lines of the products. A first homogenized data set for the entire mission duration processed with one algorithm version, namely Baseline BA, was accomplished in September 2025. We used ground-based multiwavelength-Raman-polarization lidars of PollyNET operating in the framework of the Aerosol, Clouds and Trace gases Research Infrastructure (ACTRIS) to discuss the quality of ATLID profiling products based on golden case studies. The PollyNET lidars measure the same geophysical parameter as EarthCARE, namely profiles of the particle backscatter coefficient, the particle extinction coefficient, and the particle linear depolarization ratio at 355 nm. Seven dedicated cases, for which EarthCARE and the ground-based reference system observed the same atmospheric scene, were selected, spanning several atmospheric conditions (ice clouds, high aerosol load, pristine conditions) and geographic locations (Tropical Atlantic, Europe, Central Asia, and the pristine Southern Hemisphere). Our investigations revealed that ATLID has remarkable profiling capabilities with good signal strength and high vertical resolution. The ATLID profiling product of ESA’s processing chain, A-EBD, could resolve the vertical structure of the targeted atmospheric features very well so that the A-EBD backscatter and extinction profiles (at low resolution) matched qualitatively (and mostly quantitatively) with the ground-based reference observations for most investigated atmospheric conditions. The intensive particle quantity lidar ratio is retrieved layer-wise and thus not in the same resolution as the backscatter and extinction products. It matches in many cases with the ground-based reference, but we also detected occasions when the lidar ratio in certain atmospheric regions was significantly deviating from the reference, which also then affects either the extinction or backscatter coefficient values. Especially edge effects at the transition of particle layers to clean air seem to be problematic. Concerning ATLID’s depolarization ratio, fair agreement was found for strongly scattering and depolarizing features, like ice clouds – especially during nighttime. For the aerosol regime, however, we confirm significant deviations from the ground reference and consider the depolarization ratio in Baselines BA and BB as quantitatively not reliable, especially during daytime. Thus, ATLID’s depolarization ratio of Baselines B can be used to discriminate but not to type atmospheric features.

In conclusion, we can state that ATLID’s optical profiles of Baselines B are ready for scientific exploitation keeping in mind the reported drawbacks (e.g., depolarization ratio offsets, edge effects, occasional retrieval errors, non-complete quality flags). EarthCARE data should therefore be intensively quality checked before using for scientific studies. As EarthCARE’s lifetime was recently foreseen to last for more than 10 years and algorithm development continues, such validation efforts stay important and complement other respective validation approaches.