Diurnal and nocturnal aerosol properties by AERONET sun-sky-lunar photometer measurements along four years

Results on day-to-night aerosol parameters retrieved from the Cimel CE318-T photometer, operating at Lecce_University (40 degrees 20'N, 18 degrees 6 ' E) since October 2016, are reported. The paper aims to contribute either to the studies on the CE318-T capability for monitoring aerosols in a wide range of moon's illumination and cycles and to the studies on day-to-night columnar aerosol properties by sun-sky-lunar photometry. The reliability of lunar AODs has firstly been evaluated by analysing the AOD differences calculated from the average of the last 1 h of daytime data (AOD(SS)) and the first 1 h of nocturnal data (AOD(MR)), respectively. Similarly, AOD(MS) and AOD(SR) values have been calculated from the average of the last 1 h of nocturnal data and the first 1 h of daytime data, respectively. We found that the AOD differences (AOD(SS) - AOD(MR)) and (AODMS - AODSR) were within the accuracy of solar AODs. Pairwise comparison, correlation coefficients, root mean square errors, and mean bias of AOD(SS) versus AOD(MR) and AOD(MS) versus AOD(SR) also supported the lunar data reliability. The statistical analysis of solar and corresponding lunar daily means has shown that monthly and seasonal AODs from solar and lunar photometry measurements, respectively, are characterized by similar metrics. Monthly and seasonal Angstrom exponents (angstrom) calculated at the 440-870 nm wavelength pair from daily solar and lunar AODs are also characterized by similar metrics. More specifically, pairwise comparisons by the Mann-Whitney test have shown that lunar and solar AOD and angstrom datasets, based on daily means, are not significantly different in all the months and the seasons at the p-level < 0.02. On the contrary, pairwise comparisons have shown that the Angstrom exponent differences (Delta angstrom) calculated from solar and lunar angstrom daily means are statistically different mainly in winter and autumn, likely for the high relative humidity at night-time. Therefore, Delta angstrom has been the only intensive parameter characterized by significant day-to-night changes in autumn and winter. Some case studies have been analysed/discussed to support the paper's main results and show the seasonal changes of the aerosol properties.