The role of uncertainty in nuclear winter modelling

Abstract

inherent in models of the nuclear winter phenomenon, with specific reference to the nuclear winter base term. The ‘base term’ refers to the amount of smoke injected into the atmosphere from fires ignited in a hypothetical nuclear exchange. The author examines the implications of uncertainty for estimating inputs to atmospheric circulation models used to assess the magnitude of the nuclear winter phenomenon. This is achieved through a set of experiments designed to assess the effects of variations in the spatial distribution of combustible materials on the dynamics of large urban fires. These experiments are based on an urban fire model applied to a dataset depicting the spatial distribution of combustible materials in urban areas, as derived from air photo interpretation techniques. Results indicate that the use of realistic models of the spatial distribution in place of the more common uniform and stepfunction models introduces additional uncertainty into base term estimates, due to an increase in variability in the amount of material available for ignition and an enhanced capability for fire spread in a spatially heterogeneous environment. This study shows that, paradoxically, the incorporation of more detailed information into nuclear winter base term models may introduce additional uncertainty into model outputs.