Numerical simulations of anthropogenic dust

Approximately 90% of people worldwide breathe air that contains high concentrations of particulate matter (PM) pollution. Anthropogenic dust (AD), as a crucial component of PM, can be interpreted as dust emitted by modifying or disturbing soil particles through direct (e.g., construction and driving of vehicles) and indirect (e.g., wind erosion over agricultural land, grassland, dry lakes, etc.) human activities. The compositions and properties of AD are more complex and variable than those of natural dust (ND). Current studies on dust aerosols have mostly focused on ND emissions, transport, and climate effects. However, the quantitative knowledge of the sources, characteristics, and impacts of AD is inadequate.

To comprehensively reproduce the AD emissions process, both “indirect” and “direct” AD emission were constructed to simulate AD emissions originated from diverse kinds of source regions in the study. Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) retrievals were utilized as the constraint of AD simulations. Results showed that using both indirect and direct AD emission schemes show good performance on reproducing the spatio-temporal distributions of AD at the global scale. Compared with natural dust emissions, indirect AD emissions show an indistinctive seasonal variation, with seasonal differences generally less than 0.42 μg m⁻² s⁻¹. Among indirect AD emissions, pasturelands produce more AD particles emission into the atmosphere than croplands at approximately 0.28 μg m⁻² s⁻¹, contributing 75.9% of indirect AD emissions. The developing regions emit much higher direct AD emissions than those from developed regions. In addition, AD-induced surface radiative cooling of up to -10.0 W m^-2 regionally, especially in the developing countries. The estimated global total premature mortality due to AD is 0.7 million deaths per year and is more severe in populous regions.