• Air Quality
• Atmospheric Physics
• Bayesian Inference
• Climate
• Model Evaluation

Climate is a complex, dynamic, system composed of multiple components, including the atmosphere, ocean, continental surfaces, ice sheets, and living organisms. These subsystems interact in highly nonlinear ways, influencing global climate patterns. The atmosphere is one of the most essential components of our planet, acting as a protective shield that sustains life on Earth. It is divided into distinct layers, each with unique characteristics, such as variations in temperature, pressure, and density. Among these layers, the troposphere is of primary interest, as it is where life takes place. Within the troposphere, the boundary layer plays a crucial role in regulating atmospheric dynamics by facilitating mechanical and convective turbulence, which mixes air masses throughout the atmospheric column. Here, both anthropogenic and natural emissions disperse and interact with the atmosphere, undergoing mechanical and chemical transformations that involve numerous chemical species. Since the Industrial Revolution and the beginning of standardized temperature monitoring (around 1850-1900), human activities, mainly through greenhouse gas emissions, have unequivocally contributed to global temperature increases. Geological observations further allow us to estimate temperature trends from past epochs. According to major research institutions such as the Intergovernmental Panel on Climate Change (IPCC), National Aeronautics and Space Administration (NASA), National Oceanic and Atmospheric Administration (NOAA), World Meteorological Organization (WMO), and the European Centre for Medium-Range Weather Forecasts (ECMWF), Earth’s temperature has risen by approximately 1.2°C since the pre-industrial era. These rising temperatures are linked to widespread adverse effects, including threats to food and water security, mass mortality events, biodiversity and ecosystem losses, as well as an increase in extreme weather events. Among these factors, aerosols, highly heterogeneous in composition, play a crucial role in determining air quality and, consequently, human health. Additionally, they significantly impact Earth’s energy balance due to their strong interactions with solar radiation and their role in cloud’s formation. Extreme weather events, such as cloud formation, precipitation, volcanic eruptions, dust storms, and wildfires, also cause substantial changes in atmospheric composition and dynamics. Aerosol particles, in particular, have demonstrated a cooling effect on Earth’s surface, partially counteracting the warming effects of greenhouse gases. For example, the 1991 eruption of Mount Pinatubo led to a global temperature decrease of approximately 0.5°C for 1 2 years.
This thesis focuses on the optical properties of aerosols, specifically Aerosol Optical Depth (AOD) and Ångström Exponents (ANG), due to their significance as indicators of aerosol characteristics in the atmosphere. Natural aerosols, such as desert dust, sea salt, and volcanic ash, are the primary constituents of atmospheric aerosol. This study particularly emphasizes desert dust because of its strong influence on the radiative balance and particulate matter levels. The Mediterranean region is often affected by dust outflow from Sahara and is highly sensitive to extreme weather linked to climate change like, for example, the North Atlantic Oscillation. Lampedusa Island, situated in the center of the Mediterranean Sea, provides an ideal location for studying dust events, as it is distant from major anthropogenic aerosol sources and representative of the Mediterranean basin. To analyze aerosols, multiple data sources have been used, including both ground-based observations and model simulations.
A one-year subset of observations was selected from a longer observational record at the Lampedusa Climate Observatory, which is managed by the Italian National Agency for New Technologies, Energy, and Sustainable Economic Development (ENEA). The observatory is part of NASA’s Aerosol Robotic Network (AERONET), a global network of monitoring stations dedicated to measuring aerosol optical properties and it is a national facility of the Aerosol, Clouds and Trace Gasses European Research Infrastructure (ACTRIS). These observations are compared with modeled data from the ENEA MINNI-FORAIR-IT air quality model. The core of this model, used in this work, is the Flexible Air Quality Regional Model (FARM), which simulates aerosol concentrations over Italy. Atmospheric diffusion models are fundamental tools to assess out knowledge of aerosol dynamics and interactions, but one of the main difficulties is due to the high aerosol variability in space and time. The uncertainties mainly concern the estimation of the source term and the accurate prediction of wind direction. Our first aim is to compare the AOD values obtained from the FARM model with those measured at the Lampedusa observatory, so to evaluate model performance and identify potential issues in model’s predictions, comparing results also with the Copernicus Atmosphere Monitoring Service (CAMS) data and the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model. The second objective of this study is to calibrate and test the IMPROVE (Interagency Monitoring of Protected Visual Environments) formula, originally developed by the United States Environmental Protection Agency (EPA) to quantitatively link aerosol presence to visibility, as aerosols contribute to haze and reduced visibility. The novelty of this work lies in testing new AOD parameterizations to improve aerosol forecasting, by specifically selecting desert dust events from observations carried out at the Lampedusa Observatory. New approaches are proposed to assess the validity of the IMPROVE coefficients and suggest updated ones, using a methodology that could be extended to other sources and locations. Moreover, the proposed methodologies are capable of quantifying the apportionment of aerosol species to AOD. This study represents an important case study for the Mediterranean region, with the potential for broader applications in enhancing aerosol modeling, in models like FARM, given its effectiveness in quantifying the contributions of different aerosol species to atmospheric attenuation. The goal is to improve the simulation of AOD in the ENEA modeling systems. In fact, ENEA provides daily national air quality forecasts for the next three days using the FORAIT-IT model, as well as European-level forecasts through Copernicus Services, where it is one of eleven European models in the ensemble. The insights gained from this work could also help to design air quality policies and strategies to mitigate the impact of aerosols on both human health and the environment, by separating natural contributions from anthropogenic ones.