Retrieval of aerosol properties from ground-based polarimetric sky-radiance measurements under cloudy conditions

Abstract

To study aerosol-cloud-interactions, observations in the vicinity of clouds are necessary. We have developed a new algorithm for reconstructing aerosol microphysical parameters from ground-based polarized sky radiance measurements. To enable retrievals in partly cloudy scenes, we also have implemented a new method of cloud screening. Synthetic observations have been generated with 3D Monte-Carlo radiative transfer simulations, including cuboid clouds, as well as more realistic Large-Eddy-Simulation (LES) generated cloud fields. By applying our retrieval to these synthetic datasets, we assess the accuracy of the retrieval algorithm. The 3D radiative effects of clouds on the cloudless radiance between them are investigated as a function of the distance from clouds using cuboid cloud street simulations. The LES cloud fields were used to get an estimate for the expected effects in real world scenarios, and validate the cloud-screening.For most cases, effective radius and optical thickness of the aerosol can be retrieved well, even if a significant portion (up to 70%) of the sky is covered by clouds. We find that the aerosol optical thickness is typically slightly overestimated (not more than 0.03 or 10%). The retrieval of fine mode particle effective radius is accurate to within 0.04 µm regardless of the cloud contamination. For the retrieved coarse mode effective radius the error becomes larger towards bigger particles. A positive bias in the retrieved index of refraction has been observed in most cases.Being a potential reason for the retrieval inaccuracies, the radiative effects of clouds on the adjacent clear-sky measurements have been investigated from the simulations. Total radiance is enhanced by up to 55% on average close to clouds at 550 nm, while linear polarized radiance is reduced, but only by about 25% in the same case.