The characteristic release of noble gases from an underground nuclear explosion

Highlights: • Prompt venting of radioxenon gas is not necessarily required for downwind detection at distances > 100 km. • Gas seepage at surface increases strongly with bulk permeability of explosion containment regime when > 1 darcy. • An underground nuclear explosion may be more readily detected by standoff atmospheric sampling than by on-site radiation surveys. • Tracking atmospheric plumes during an on-site inspection may allow more rapid location of an underground test site. • Nighttime sampling of atmospheric gases can increase the distance of detection. - Abstract: Prompt release of gases at the ground surface resulting from explosively propagated vents or large operational releases has typically been considered to be the only mode of transport for detonation gases from an underground nuclear explosion (UNE) giving rise to detectable levels of radioxenon gases in downwind atmospheric samples captured at distances exceeding 100 km. Using a model for thermally and barometrically driven post-detonation transport across the broad surface of a simulated UNE site, we show in conjunction with the results of an atmospheric tracer-release experiment that even deep, well-contained UNEs, without prompt vents or leaks, are potentially detectable tens of kilometers downwind with current technology; distances that are significant for localizing the source of detected atmospheric signals during on-site monitoring or inspection. For a given yield, the bulk permeability of the UNE site and to a lesser extent the depth of detonation appear to be the primary source-term parameters controlling the distance of detection from the detonation point. We find for test-site bulk permeabilities exceeding 1 darcy (10⁻¹² m²) that broad-area surface fluxes of radioxenon gas exhibit exponential dependence on permeability resulting in order-of-magnitude enhancements of surface flux for changes in permeability of only a darcy. Simulations of subsurface transport assuming a canonical detonation-depth-versus-nuclear-yield relationship generally resulted in larger atmospheric signals for shallower, lower-yield explosions allowing downwind detection at distances greater than 1000 km. Additionally, atmospheric simulations suggest that the lowest atmospheric boundary layer heights, such as occur at night, produced concentrations above minimum detectable levels at the greatest distances downwind.