Abstract ID: 3.12413 | Accepted as Talk | Poster | TBA | TBA

Influencing the radiation budget of the atmosphere and air quality, aerosol particles are crucial in climate change research and public health. Aerosols can have a wide range of optical and microphysical properties with high temporal and spatial variability, which poses a challenge in comprehensively characterizing them. The aerosol vertical distribution is relevant for several reasons: as a priori information for satellite retrievals, for constraining the impact on radiative processes and cloud formation and for particle concentration at the surface.
Lidar instruments are well established tools for retrieving aerosol profiles. Passive ground-based remote sensing measurements also provide information on aerosol profiles, but there is no established retrieval method yet. With their highest sensitivity being at low altitudes, they could complement the lidar method, which is limited to altitudes above 500m to 1000m. The Pandora spectrometer system and the Cimel photometer routinely perform direct sun and sky measurements within the Pandonia Global Network (PGN) and the Aerosol Robotic Network (AERONET). Trace gas profiles of NO2, HCHO and H20 are already retrieved based on so-called Multi-Axis Differential Optical Absorption Spectroscopy (MAX-DOAS) observations, i.e. spectral sky radiance measurements at various elevation angles (at a constant azimuth angle) and applying a spectral fitting method.
Aerosol profiles are more complex to retrieve because their impact on sky radiance has no spectral signature and depends on their optical properties, which may vary widely between different types.
Here we report on first results of aerosol profile retrievals from passive ground-based remote sensing for Innsbruck, Austria. In this study we analyse two approaches. First, we use a parametrized approach based on relative sky radiance and retrieved absolute slant column densities of the oxygen dimer (O2O2) from Pandora. AOD per layer is calculated based on the comparison with look-up tables of a pure Rayleigh atmosphere and scaled by total AOD. This approach would facilitate the integration as near-real-time PGN product.
Second, we apply the optimal estimation technique of a modified version of the GRASP algorithm, fitting the measurements of sky radiance from the AERONET and slant columns from the PGN to a radiative transfer model.