Aerosol optical depth and water vapor variability assessed through autocorrelation analysis

Abstract

Numerous studies globally have centered on atmospheric air pollution due to its profound health and climate effects. NASA’s AERONET (National Aeronautics and Space Administration - AErosol RObotic NETwork) network has been one of the world’s leading tools for accessing the physical properties of atmospheric aerosols from various sources, mainly anthropogenic ones. This study proposes a new approach to evaluate the Aerosol Optical Depth (AOD) and precipitable water vapor (PWV) seasonality and the influence of short-term perturbations, such as the presence of local and regional aerosol sources or meteorological events, based on the temporal autocorrelation function (ACF). We introduce the adimensional seasonal assessment autocorrelation function, $\Delta _{{{\text{ACF,k}}}}$, as a parameter to quantify the influence of the short-term perturbation, and we use its average, $\langle \Delta _{{{\text{ACF,k}}}} \rangle$, as a proxy for seasonality loss. The smaller $\langle \Delta _{{{\text{ACF,k}}}} \rangle$, the lower the influence of high-frequency perturbations on seasonality. Nine AERONET network sites in South America with different environmental characteristics were evaluated. The selected sites were São Paulo, Rio Branco, Manaus, ATTO (Amazon Tall Tower Observatory), Alta Floresta, Ji-Paraná, Cuiabá, Arica, and La Paz. The results showed that sites with less local anthropogenic aerosol sources acting as short-term perturbations had pronounced AOD seasonality and a linear relationship between the ACF functions of AOD, PWV, and the simulated direct solar radiation. As local anthropogenic sources become more prominent, the AOD ACF is attenuated and has less amplitude in seasonal oscillations. In addition, the relationship between AOD and PWV ACF becomes more attenuated. Buenos Aires has shown to be the most affected site, with $\langle \Delta _{{{\text{ACF,AOD}}}} \rangle$ of 0.47, followed by São Paulo and La Paz. The areas in the Amazonian deforestation arc had relatively close average $\Delta _{{{\text{ACF,AOD}}}}$, with Alta Floresta representing the most influenced by short-term perturbations. Central Amazonian sites had the lowest $\Delta _{{{\text{ACF,AOD}}}}$ averages, of about 0.25, which means that constant local anthropogenic sources do not dominate the AOD seasonality and that the wet deposition still plays an essential role in regulating the aerosol sources in the atmosphere. In contrast, the behavior of $\langle \Delta _{{{\text{ACF,PWV}}}} \rangle$ in the Amazon region varies mainly due to meteorological influences, with the highest values observed in the central region, likely related to the high amount of water vapor in the atmosphere, and more pronounced seasonality near deforestation arcs and major cities. The proposed method eliminates the need for a reference site when comparing seasonalities of different time series, enabling valid comparisons across different areas without a comparative reference point. The method can be further applied to other atmospheric time series, including greenhouse gases.