Multi-instrumental approach for air quality monitoring: characterization and distribution of greenhouse gases and atmospheric metal deposition in the Greve River Basin (Chianti territory, Central Italy).

Monitoring networks, able to effectively provide high-frequency geochemical data for characterizing the geochemical behavior of the main greenhouse gases (i.e., CO₂ and CH₄) and pollutants (e.g., heavy metals) are crucial tools for the assessment of air quality and its role in climate changes. However, the provision of measurement stations dedicated to monitor gas species and particulate in polluted areas is complicated by the high cost of their set-up and maintenance. In the last decade, traditional instruments have tentatively been coupled with low-cost sensors for improving spatial coverage and temporal resolution of air quality surveys. The main concerns of this new approach regard the in-field accuracy of the low-cost sensors, being significantly dependent on: (i) cross-sensitivities to other atmospheric pollutants, (ii) environmental parameters (e.g., relative humidity and temperature), and (iii) detector signal degradation over time.

This study presents the results of a geochemical survey carried out in the Greve River Basin (Chianti territory, Central Italy) from May to September 2022 by adopting two measuring strategies: (i) deployment of a mobile station, along predefined transepts within the Greve valley, equipped with a Picarro G2201-i analyzer to measure CO₂ and CH₄ concentrations and δ¹³C-CO₂ and δ¹³C-CH₄ values (‰ vs. V-PDB) by Wavelength-Scanned Cavity Ring-Down Spectroscopy (WS-CRDS); (ii) continuous monitoring, at five fixed sites positioned at different altitudes, of CO₂ and CH₄ concentrations through prototyped low-cost stations, coupled with atmospheric deposition and rain samplers to collect particulate samples for chemical lab analysis. The low-cost monitoring stations housed (i) a non-dispersive infrared (NDIR) sensor for CO₂ concentrations, (ii) a solid-state metal oxide sensor (MOS) for CH₄ concentrations, (iii) a laser light scattering sensor (LSPs) for PM_2.5 and PM₁₀ concentrations, and (iv) a sensor for temperature and relative humidity in the air. The CO₂ and CH₄ sensors have been calibrated in-field based on parallel measurements with the Picarro G2201-i and elaborating the calibration data with the Random Forest machine learning-based algorithm.

The measurements carried out along the transepts showed that the downstream areas next to the metropolitan city of Florence were affected by the highest concentrations of CO₂ and CH₄, marked by isotopic signatures revealing a clear anthropogenic origin, mainly ascribed to vehicular traffic. The distribution of these carbon species reflected the evolution of the atmospheric boundary layer, displaying higher concentrations during the early morning, when gas accumulation occurred due to stable atmospheric conditions, and lower concentrations during daytime when the heating of the surface favored the dilution of air pollutants due to the establishment of convective turbulence. These observations were confirmed by the network of low-cost stations, which allowed to simultaneously monitor the distribution of the atmospheric pollutants at different altitudes in the valley. The distribution of particulate was consistent with that of the gaseous species, and the main sources were clearly distinguished based on the chemical composition of the atmospheric deposition in the collection sites. The promising results from the present study could result in an affordable approach to effectively improve air quality monitoring strategies and support data-driven policy actions to reduce carbon emissions.