Thesis title: Advanced Retrieval Algorithms for Atmospheric Aerosol Monitoring through Lidar-Ceilometer Networks and Integration with Other Remote Sensing and Modelling Techniques
Atmospheric aerosols significantly impact Earth's climate and air quality. Their diverse emission sources and atmospheric processes result in considerable spatiotemporal variability that requires continuous monitoring. Active remote sensing through Light Detection and Ranging (lidar) enables vertical profiling of atmospheric particles, which is essential for evaluating aerosol radiative effects, transport dynamics, formation processes, and aerosol-cloud interactions. In recent years, Automated Lidar Ceilometers (ALCs) have become valuable tools for aerosol profiling, offering continuous, unattended operation that has driven the development of extensive networks worldwide. However, extracting quantitative information from ALC measurements presents significant challenges, and most networks currently lack comprehensive data processing to derive the geophysical quantities needed by atmospheric research and monitoring communities. Furthermore, the potential for ALC integration with other atmospheric monitoring systems is presently undervalued.
This research addresses these challenges by developing advanced algorithms for ALC networks that enable continuous, quantitative aerosol profiling and integration with other monitoring platforms. This work, conducted within the Italian Automated Lidar-Ceilometer Network (ALICENET) and contributing to EU research initiatives, produced four main outcomes (partly described in five publications, three of which are first-authored: Bellini et al., AMT, 2024; Bellini et al., Rem. Sens., 2025a; Bellini et al., in preparation, 2025b). First, it developed innovative retrieval methodologies to extract quantitative atmospheric/aerosol data from ALC observations. Refined data harmonisation procedures improved near-surface monitoring capabilities and calibration accuracy. The retrieval of aerosol properties implemented a model-based approach that bypasses the need for ancillary measurements, providing aerosol extinction and volume/mass concentration profiles. A novel ALADIN (Aerosol LAyer DetectIoN) algorithm was designed and implemented to provide continuous characterisation of aerosol stratifications in the troposphere, identifying mixed, continuous, and elevated aerosol layers. Extensive validation of the algorithms demonstrated retrieval accuracy with 30% agreement for aerosol optical properties compared to research-grade measurements (e.g., from the NASA Aerosol Robotic Network) and 40% agreement for mass concentrations against in-situ observations.
Second, the research established a comprehensive processing framework suitable for operational use within networks. This framework incorporates quality control procedures at each processing stage and produces outputs ranging from basic near-real-time monitoring (Levels 1-2) to advanced aerosol retrievals (Level 3) for research applications.
Third, the methodologies were successfully applied across various domains, demonstrating effective integration with other remote sensing systems and models. The processing chain enabled real-time tracking of regional-to-intercontinental aerosol transport events, including Saharan dust intrusions and Canadian wildfire plumes, while supporting weather forecasting through improved representation of aerosol-radiation interactions. The profiling approaches enhanced air quality monitoring in urban areas (Paris, Rome, and the Po Valley), revealing multiple atmospheric processes impacting surface particulate matter concentrations. Application of the new tools over long-term ALC records across Italian monitoring sites comprehensively characterised aerosol vertical distributions and dynamics from sub-hourly to seasonal timescales. The analysis revealed marked latitudinal and seasonal gradients: the northern Alpine site exhibited median particulate matter concentrations below 15 μg m⁻³ throughout the column, while urban and coastal sites in central and southern Italy showed values exceeding 40 μg m⁻³ in the lower troposphere. During winter months, 70-80% of aerosol optical depth was concentrated within the first kilometre, while in summer, 30-60% occurred above 2 km altitude. Elevated aerosol layers, detected during 40% of summer days, significantly impacted surface air quality for 10-40% of these days. Comparison with state-of-the-art model reanalyses found model underestimation of particulate matter by 5-15 μg m⁻³ in the 1-3 km altitude range and 10-30% agreement in the representation of mixed layer dynamics.
Fourth, the work favoured and supported the expansion of ALC monitoring capabilities in Italy, filling an observational gap at national and EU scales.
Overall, this research advances the Italian atmospheric aerosol monitoring capability. Its planned full implementation within ALICENET allows continuous profiling of aerosol optical properties, particulate matter vertical distributions, and mixed/elevated layer evolution for air quality assessments, bridging the gap between research-oriented and operational applications.