Date: Wednesday 12 December 2012
The effect of global aerosols on clouds is one of the largest uncertainties in the radiative forcing on the climate1. Cloud condensation nuclei (CCN) are large soluble aerosol in the atmosphere onto which clouds form and are a key quantity in understanding the aerosol-climate effect. Global aerosol models2 are used to simulate the global distribution of aerosol from emission or nucleation to CCN but these models come with a degree of uncertainty. We perform global variance-based sensitivity analysis (SA) via emulation to identify the leading causes of parametric uncertainty in CCN3,4. Performing SA on every model grid box throughout the year 2008 we can identify the processes leading to uncertainty in model predicted CCN. The same technique is used to show that the most important uncertainties in model predicted CCN are not the same as those in indirect forcing due to changes between the pre-industrial and present day aerosol processes.