Type-segregated aerosol effects on regional monsoon activity: A study using ground-based experiments and model simulations

Abstract

Classification of observed aerosols into key types [e.g., clean-maritime (CM), desert-dust (DD), urban-industrial/biomass-burning (UI/BB), black carbon (BC), organic carbon (OC) and mixed-type aerosols (MA)] would facilitate to infer aerosol sources, effects, and feedback mechanisms, not only to improve the accuracy of satellite retrievals but also to quantify the assessment of aerosol radiative impacts on climate. In this paper, we report the results of a study conducted in this direction, employing a Cimel Sun-sky radiometer at the Indian Institute of Tropical Meteorology (IITM), Pune, India during 2008 and 2009, which represent two successive contrasting monsoon years. The study provided an observational evidence to show that the local sources are subject to heavy loading of absorbing aerosols (dust and black carbon), with strong seasonality closely linked to the monsoon annual rainfall cycle over Pune, a tropical urban station in India. The results revealed the absence of CM aerosols in the pre-monsoon as well as in the monsoon seasons of 2009 as opposed to 2008. Higher loading of dust aerosols is observed in the pre-monsoon and monsoon seasons of 2009; majority may be coated with fine BC aerosols from local emissions, leading to reduction in regional rainfall. Further, significant decrease in coarse-mode AOD and presence of carbonaceous aerosols, affecting the aerosol–cloud interaction and monsoon-rain processes via microphysics and dynamics, is considered responsible for the reduction in rainfall during 2009. Additionally, we discuss how optical depth, contributed by different types of aerosols, influences the distribution of monsoon rainfall over an urban region using the Monitoring Atmospheric Composition and Climate (MACC) aerosol reanalysis. Furthermore, predictions of the Dust REgional Atmospheric Model (DREAM) simulations combined with HYSPLIT (HYbrid Single Particle Lagrangian Integrated Trajectory) cluster model are also discussed in support of the observed features.