Ice nucleating ability of mineral particles from subtropical South American deserts

Abstract

Mineral aerosols are one of the most important ice nucleating particles (INPs) because their efficiency in nucleating ice, wide transport and largest mass contribution to particulate matter in the atmosphere. They are sourced from the arid regions of the world. In this context, this work evaluates the INP potential of fourteen topsoil samples collected from subtropical South American deserts, the major source of mineral aerosols in South America, in the immersion freezing mode. Samples were obtained from three distinct regions located in the South American Arid Diagonal and recognized as potential dust source areas: the Puna-Altiplano Plateau in the north, the central-west of Argentina, and Patagonia in the south. In general, results reveal that samples from the Puna-Altiplano and Patagonia regions, and the central-west of Argentina region exhibit the highest and lowest INP abilities, respectively. The active sites per unit surface area for a given temperature were calculated and compared with previously reported values. The results demonstrate that soil mineral particles from the region of study exhibit ice nucleating abilities comparable to the inorganic fraction of agricultural soils of central Argentina. No direct relationship was identified between INP ability and the major minerals observed in the samples. This study is the first to analyze the ice nucleation properties of soil samples collected along the South American Arid Diagonal and one of the few in South America. Since the analyzed topsoil particles were collected from potential dust source regions, this work contributes to understanding the role of aerosols in initiating atmospheric ice formation, providing valuable data for empirical parameterizations. This could contribute to the improvement in the performance of climate models, as the obtained results suggest that the underestimation of coarse and super-coarse aerosols at altitudes relevant for cloud formation may lead to underestimations in INP concentrations, particularly in regions near to the emission sources.