Retrieving Aerosol Properties from Lidar-Ceilometers in Dust Conditions Using GRASP and Monte Carlo Methods

The atmosphere contains not only gases but also tiny solid and liquid particles called aerosols, which play a crucial role in air quality, weather, and climate. Ground-based remote-sensing instruments, known as Automated Lidar-Ceilometers (ALCs), measure the presence of these particles along the vertical profile by emitting laser pulses and recording the light scattered back to the Earth’s surface. However, while ALCs measure light, aerosol geophysical relevant quantities are often required in atmospheric studies and applications. For instance, air quality regulations are typically based on estimates of particle mass within a volume of air. To address this, conversion functions based on optical models have been developed. However, a major limitation of the current methodologies is the assumption that aerosol particles are mostly spherical as these methods were developed to represent continental aerosol conditions. In presence of desert dust particles, which have non-spherical shapes, using such models can lead to inaccurate retrievals of geophysical properties, for example resulting in underestimation of aerosol mass. New generations of polarization-sensitive ALCs provide crucial information about particle shape and operate at different wavelengths, making it necessary to develop updated conversion functions. This thesis aims to retrieve key aerosol properties, such as extinction coefficients, surface area, and volume concentration, from polarization-sensitive ALC measurements at 355, 532, 910, and 1064 nm. The method is based on a large dataset of scattering simulations, with optical and microphysical properties selected via a Monte Carlo approach. The main innovation is the use of a mixture of spheroids to simulate the coarse aerosol fraction, which better approximates the scattering properties of non-spherical particles. Simulations were performed using the spheroid module of the GRASP (Generalized Retrieval of Atmosphere and Surface Properties) framework, from which mean functional relationships were derived linking the aerosol backscatter coefficient and the particle depolarization ratio to the target variables. Results show an average increase in extinction coefficient between 9% and 17% and in volume concentration between 59% and 79% compared to the standard method. This led to improved agreement of aerosol optical depth (AOD) with AERONET and of mass concentrations with Optical Particle Counter (OPC) measurements. The updated method significantly improves performance during dust events, which the previous method tended to underestimate. This work can be directly implemented in ALICEnet, the Italian Automated Lidar-Ceilometer Network. This will enhance the quality of aerosol retrievals from both standard and polarization-sensitive ALCs, improving aerosol monitoring and characterization across Italy and providing valuable data for studies on radiative transfer, atmospheric chemistry, and air quality.