Global assessment of aerosol radiative effects: New insights from observations, reanalysis, and climate models

Aerosols continue to be the most uncertain forcing factor in quantifying the present-day radiative forcing. An integrated analysis of crucial aerosol optical and radiative properties across the globe using multi-source (ground-based, satellite, reanalysis, and climate model) data and quantifying the biases on regional and temporal scales can substantially reduce this uncertainty. In a first-of-its-kind study, a comprehensive investigation using AErosol RObotic NETwork (AERONET), MODerate resolution Imaging Spectroradiometer (MODIS), Ozone Monitoring Instrument (OMI), Clouds and the Earth's Radiant Energy System (CERES), Modern-Era Retrospective Analysis for Research and Applications-2 (MERRA-2), and Coupled Model Intercomparison Project Phase 6 (CMIP6) datasets, revealed that aerosol direct radiative effect (DRE) is highest over South Asia and lowest over Australia, followed by North America and Europe. AERONET retrieved aerosol DRE at surface (DRE_SFC: ∼ - 70 Wm^-2), top-of-atmosphere cooling (DRE_TOA: ∼ - 30 Wm^-2), and atmospheric heating (DRE_ATM: ∼40 Wm^-2; HR: ∼0.80 Kday^-1) are strongest over South Asia - I with significant spatiotemporal variations. These values of DREs arise due to high aerosol optical depth (AOD: ∼0.57) and low single scattering albedo (SSA: ∼0.91). In contrast, AOD and DRE (absolute) in North America, Europe, and Australia are ∼2-4 times lower than in Asia, with less spatiotemporal variability. Notably, low SSA enhances both atmospheric heating and surface cooling efficiencies over South Asia - I and biomass-burning regions. High underestimation of MERRA-2 AOD in high AOD conditions leads to high underestimations in MERRA-2 DREs (in absolute terms). Underestimations in MERRA-2 DREs are pronounced over Asia (∼25 %), with large biases over South Asia (∼33 %). Taylor diagram analysis and collocated validation reveal that MERRA-2 outperforms CERES in reproducing AERONET DRE globally; however, both datasets exhibit substantial biases over Asia. On an annual scale, CMIP6 multi-model mean underestimates AOD in South and Southeast Asia (factor of ∼1.4), and the differences between CMIP6 aerosol direct radiative forcing (DRF) and AERONET DRE at each level (top-of-atmosphere, atmosphere: >10 Wm^-2; surface: >20Wm^-2) are higher over most regions in Asia. This study provides crucial global insights into aerosol direct radiative effects by utilizing multi-platform datasets, quantifying the regional and seasonal biases, which are essential to fine-tune and improve the aerosol properties and processes in regional and global climate models for accurately assessing the aerosol-climate interactions.