A novel method to quantify the uncertainty contribution of aerosol–radiation interaction factors

Abstract

The IPCC's assessment report shows that the radiative forcing of aerosol–radiation interactions still involves significant uncertainty. The commonly used method for factor uncertainty estimation is the one-at-a-time (OAT) method, which evaluates factor sensitivity by controlling the change in a single variable while keeping others constant. The outcomes from the OAT method require high data quality to ensure accuracy, and the results are only valid near the selected constant. This study proposes a new method called the Constrained Parameter (CP) method to quantify the uncertainty contribution of factors in a multi-factor system. This method constrains the uncertainty of a single factor between two Monte Carlo simulations and evaluates its sensitivity by analyzing how this change affects output uncertainty. The most significant advantage of the CP method is that it can be applied to any data distribution, and its results can reflect the overall data characteristics. The proportion of factor interactions in the factor uncertainty contributions can be obtained by comparing the results calculated by the CP method and the OAT method. As an application of the CP method, it performs a detailed analysis of aerosol–radiation interaction factors' uncertainty contributions. The top three most sensitive factors are the complex refractive index of aerosol shell materials, light-absorbing carbon parameters, and Mie theory parameters. Due to their high sensitivity and low observational precision, these factors represent significant sources of uncertainty in aerosol–radiation interactions. These factors need to be prioritized for operational observation programs and model parameter inputs.