Variability in Biomass Burning Emissions Weakens Aerosol Forcing Due To Nonlinear Aerosol‐Cloud Interactions

Abstract

Atmospheric aerosols are a critical component of the climate system, but the complex processes governing their production, deposition, and interactions with clouds are difficult to observe and model. Uncertainty in the aerosol forcing is one of the greatest challenges for understanding historical climate change and projecting near-future climate evolution (Forster et al., 2021;Kiehl, 2007).Previous research on aerosol radiative forcing has focused on the effect of secular change in aerosol emissions, with little consideration of the impact of shorter timescale variability in the emissions. For example, the fifth Coupled Model Intercomparison Project (CMIP5; Taylor et al., 2012) historical and future simulations use biomass burning (BB) emissions estimates that are smooth temporally compared to real-world emissions, particularly on inter-and sub-annual time scales. Real-world BB emissions in the extratropics occur episodically and stochastically, and may depend on weather conditions (precipitation, drought, lightning) or human activity (agricultural burning, forest clearing, arson) (Lamarque et al., 2010;van der Werf et al., 2017).To incorporate more realistic aerosol emissions variability, the latest CMIP (sixth phase; CMIP6; Eyring et al., 2016) includes BB emissions estimates derived from satellite observations for historical simulations from 1997 to 2014 (Figure 1a; van Marle et al., 2017). Historical CMIP6 BB emissions in this time period have much higher temporal variability than those used in previous model intercomparison efforts (e.g., the CMIP5 historical simulations). However, the BB emissions used for CMIP6 prior to 1997 (before satellite measurement capability) are similar to the CMIP5 inventories, with weak temporal variability (Figure 1a black line; Lamarque et al., 2010; van Marle et al., 2017). Recent analyses in the Community Earth System Model version 2 (CESM2; Danabasoglu et al., 2020) have estimated the climate effect of this change in BB emissions variability by comparing simulation scenarios