0.6 for various productive ecosystems)./p> 250 m), we consider only the bin that has the closest intersection with the DEM. Next, we corrected molecular and aerosol contributions using Eq. (2) and obtained final surface-based LSR estimates (γ). Eq. (2) shows how molecular and aerosol corrections were applied using the Rayleigh optical depth (ODRay; see Supplementary Material, S1) and Aerosol Optical Depth (AOD; see Supplementary Material, S2), respectively. In theory, LSR over land can be converted into BRDF using a 2π correction factor, while such conversion has been applied mostly to the nadir looking CALIOP with the incidence angle close to 3°6. The same approach might not be applicable for highly non-nadir lidars like Aeolus because one does not take into account the angles of incidence and refraction. Over water surfaces where much more complex interaction between specular, whitecap and subsurface reflectance components may occur the LSR and BRDF comparison is more complicated17./p> 1.0 We calculated AOD using the Aeolus Profile Processor Algorithm (AEL-PRO), which relies on the optimal estimation and forward modelling inversion procedure. In short, AEL_PRO retrieves the lidar-to-backscattering ratio profile by using only the pure Rayleigh and Mie attenuated backscatter values as input, thereby yielding accurate extinction coefficient profiles22. The output profiles of the retrieved state vector, including aerosol/cloud extinction coefficients, were utilized in this work to estimate AOD. Moreover, since AEL_PRO can categorize atmospheric features (water-cloud, ice-cloud, aerosol, clean sky, etc.), we applied the most stringent filtering strategy by excluding any LSR observations potentially contaminated by ice cloud and water cloud presence. In simple words, we used AEL_PRO to keep only the high quality LSR observations without clouds, where the surface signal was not attenuated. This filtering was performed at the highest measurement resolution of Aeolus. For AEL_PRO details, see supplementary material S2 and Donovan et al.22 work. We subsequently calculated monthly averages of the LSR with the corresponding standard deviations on a 2.5° × 2.5° grid in the first full yearly (or seasonal) cycle of Aeolus observations (September 2018–August 2019). In the study period, the monthly averaging of millions of observations yields an abundant quasi-global coverage by LSR observations. Minor temporal data gaps were present only during some days in January and February 2019, when Aeolus experienced a system failure. We also did not use any data from June 2019 due to the change from the Flight model-A laser, FM-A, to the second laser, FM-B, period14 to avoid any negative effects of the shift of the regime during the same month. Each step of the LSR calculation is illustrated in detail in the supplementary material (SI Fig. S1). Note that although we thoroughly addressed all potentially malignant effects for LSR estimation, some limitations stem directly from the Aeolus setup. Most importantly, the emitted lidar pulse is circularly polarized; however, the Aeolus receiver is only detecting the co-polar component, which could lead to discrepancies in the LSR estimations. This limitation is inherent as Aeolus does not have a depolarization channel. Future LSR estimations from Aeolus may be revisited when the EarthCARE mission is launched, which includes a linearly polarized lidar instrument at the same wavelength with a depolarization channel. This will allow an estimation of both the circular and co-polar components of depolarization, which can then be compared to the Aeolus LSR estimates in a retrospective analysis./p>