Revealing dominant patterns of aerosol regimes in the lower troposphere and their evolution from preindustrial times to the future in global climate model simulations

• Published in: Atmospheric chemistry and physics 2024-11, Vol.24 (22), p.12727-12747
• Main Authors: Li, Jingmin, Righi, Mattia, Hendricks, Johannes, Beer, Christof G, Burkhardt, Ulrike, Schmidt, Anja
• Format: Article
• Language: English
• Subjects: Aerosol properties; Aerosols; Air pollution; Air pollution measurements; Algorithms; Analysis; Anthropogenic factors; Atmospheric aerosols; Atmospheric chemistry; Classification; Climate change; Climate effects; Climate models; Climatic analysis; Climatic changes; Cloud properties; Clustering; Developed countries; Developing countries; Emission analysis; Emissions; Evolution; Global aerosols; Global climate; Global climate models; Human influences; LDCs; Lower troposphere; Machine learning; Marine aerosols; Outdoor air quality; Pollution control; Radiation; Radiation-cloud interactions; Radiative forcing; Regional development; Simulation; Simulation methods; Spatial distribution; Spatial variability; Spatial variations; Troposphere
• ISSN: 1680-7324; 1680-7316; 1680-7324
• DOI: 10.5194/acp-24-12727-2024

Aerosols play an important role in the Earth system, but their impact on cloud properties and the resulting radiative forcing of climate remains highly uncertain. The large temporal and spatial variability of a number of aerosol properties and the choice of different “preindustrial” reference years prevent a concise understanding of their impacts on clouds and radiation. In this study, we characterize the spatial patterns and long-term evolution of lower tropospheric aerosols (in terms of regimes) by clustering multiple instead of single aerosol properties from preindustrial times to the year 2050 under three different Shared Socioeconomic Pathway (SSP) scenarios. The clustering is based on a combination of statistic-based machine learning algorithms and output from emissions-driven global aerosol model simulations, which do not consider the effects of climate change. Our analysis suggests that in comparison with the present-day case, lower tropospheric aerosol regimes during preindustrial times are mostly represented by regimes of comparatively clean conditions, where marked differences between the years 1750 and 1850 emerge due to the growing influence of agriculture and other anthropogenic activities in 1850. Key aspects of the spatial distribution and extent of the aerosol regimes identified in year 2050 differ compared to preindustrial and present-day conditions, with significant variations resulting from the emission scenario investigated. In 2050, the low-emission SSP1-1.9 scenario is the only scenario where the spatial distribution and extent of the aerosol regimes very closely resemble preindustrial conditions, where the similarity is greater compared to 1850 than 1750. The aerosol regimes for 2050 under SSP3-7.0 closely resemble present-day conditions, but there are some notable regional differences: developed countries tend to shift towards cleaner conditions in future, while the opposite is the case for developing countries. The aerosol regimes for 2050 under SSP2-4.5 represent an intermediate stage between preindustrial times and present-day conditions. Further analysis indicates a north–south difference in the clean background regime during preindustrial times and close resemblance of preindustrial aerosol conditions in the marine regime to present-day conditions in the Southern Hemispheric ocean. Not considering the effects of climate change is expected to cause uncertainties in the size and extent of the identified aerosol regimes but not t