Constraining the Simulated Radiative Effects of Biomass Burning in Southern Africa

Seasonal biomass burning fires are common in several regions around the world including South America, Australia, Southern Africa, and others. On a global scale, the region of southern Africa is important because it releases the largest portion of the planet's biomass burning aerosols into the atmosphere, and because the prevailing winds carry the smoke over the south Atlantic ocean where they interact with clouds. This area of the Atlantic contains the world's largest semi-permanent stratocumulus cloud deck. Aerosol interactions affect cloud albedo and may also influence the transition from the stratocummulus to cumulus cloud state. In this project, we constrain the uncertainty in the radiative effects of the Biomass Burning events that occur seasonally in Southern Africa that we simulate in a climate model by considering the uncertainty in the input parameters to that model. We hypothesize that we can parameterize a large fraction of the uncertainty in smoke radiative effects by introducing or varying 12 uncertain parameters in the model. Some parameters scale emissions, both of smoke and of natural aerosols and precursors, and others scale parameters in uncertain processes such as the vertical velocities used to activate cloud droplets. Since satellites are not capable of measuring radiative effects, we start by comparing simulated aerosol optical depth (AOD) with satellite measured AOD to rule out combinations of parameters that produce implausible results. Prior work (e.g. Johnson et al, Atmos Chem Phys 2020) has shown AOD is capable of constraining radiative forcing. We calculate the fire radiative effects by comparing simulations with and without smoke emissions. In future we will use combinations of several satellite variables including AOD to develop a stronger constraint on the radiative effect of these fires. To constrain our uncertain parameters, we built a new perturbed parameter ensemble consisting of 121 runs of the atmosphere-only UK Earth System Model spanning our space of 12 parameters. We use this as training data to build surrogate models. We simulate only the 2017 biomass burning season, and write data out from the simulations at three-hourly time resolution, which allows us to understand the transport of individual smoke plumes. The surrogate models are trained using Gaussian Process Regression. To rule out implausible parameter combinations that yield simulated AODs that are inconsistent with the observations, we use statistical methods developed in Carzon et al (Environmental Data Science, 2023). These allow us to quantify uncertainty in the parameters at a high level of statistical confidence, improving on previous related work using the method of 'history matching' (whose theoretical guarantees are not clearly understood). We build on the statistical methods of Carzon et al. and aim to comprehensively validate these methods for our setting. We will present the methods, the constraints on parameters we have achieved to date, and the implications for the smoke radiative effects.