Investigating the Radiative Forcing from Volcanic Eruptions Using Statistical Emulation

Abstract:

The magnitude of the radiative forcing of a volcanic eruption is dependent on several eruption source parameters such as the mass of sulfur dioxide (SO₂) emitted, the eruption column height, the eruption latitude, eruption season and the phase of the Quasi-Biennial Oscillation. Global climate models with prognostic aerosol and chemistry schemes can be used to investigate the importance of each parameter in determining the volcanic forcing. However, the range of parameter space that can be explored is restricted due to the computational expense of these simulations. Here we use Gaussian process emulators to estimate the impact of eruptions over a range of SO₂ injection magnitudes, injection heights and latitudes based on a relatively small number of model simulations. We show how common metrics of the volcanic forcing efficacy - global-mean stratospheric Aerosol Optical Depth (AOD) at 550 nm, net radiative forcing and sulfate aerosol e-folding time, depend on the three parameters. For AOD and radiative forcing, SO₂ is the most important parameter, but latitude is the most important parameter in determining the sulfate e-folding time. Most importantly, our emulators can be used to predict the AOD, net radiative forcing and sulfate e-folding time of any eruption within our parameter space that has not been explicitly simulated.