Large and sensitive simulated radiative effects of smoke in the south-east Atlantic

Abstract:

Smoke aerosols exert large and complex radiative effects on Earth’s climate¹. The direct radiative effects of biomass burning aerosols are highly uncertain. As well as these direct radiative effects, biomass burning aerosols can strongly affect cloud cover and albedo both by heating clouds (the semi-direct effect) and by acting as cloud condensation nuclei. In our simulations, about 20% of cloud condensation nuclei at the altitude of low clouds globally originate from biomass burning. Semi-direct effects are particularly difficult to quantify but may lead to radiative effects of tens of Wm^-2 over large areas of Earth’s ocean surface².

In this talk we will use an atmosphere-only configuration of the HadGEM3-GC3.1 climate model (as submitted to CMIP6) to illustrate how the simulated direct, indirect and semi-direct radiative effects of biomass burning aerosol depend on aerosol and cloud properties, as well as on the details of how a simulation is performed and how the radiative effects are calculated. For example, we consider the simulated total radiative effects of smoke over the whole of the south-east Atlantic over three weeks of August 2017. We find these can differ by a factor of almost two between free-running simulations and simulations in which horizontal winds are nudged to reanalysis in the free troposphere, despite only modest differences in the simulated meteorology.

We will focus on the south-east Atlantic region, building on earlier work². This region is the destination for much of the smoke from southern Africa, the world’s largest source of biomass burning aerosol. The ORACLES, CLARIFY and LASIC field campaigns³ are used together with satellite retrievals to constrain simulated aerosol and cloud microphysical properties in the region. Of particular interest are changes in aerosol composition and optical properties as the aerosols advect across the ocean and are subject to photochemical ageing and in-cloud processing^4,5.

¹Myrhe et al, Atmos Chem Phys. 2013; ²Gordon et al, Atmos Chem Phys. 2018; Zuidema et al, Bull. Am. Met. Soc. 2016, ⁴Wu et al, Atmos Chem Phys 2020; ⁵Shinozuka et al, Atmos Chem Phys 2020