Modeling Collapse Chimney and Spall Zone Settlement as a Source of Post-shot Subsidence Detected by Synthetic Aperture Radar Interferometry

Ground surface subsidence resulting from the March 1992 JUNCTION underground nuclear test at the Nevada Test Site (NTS) imaged by satellite synthetic aperture radar interferometry (InSAR) wholly occurred during a period of several months after the shot (Vincent et al., 1999) and after the main cavity collapse event. A significant portion of the subsidence associated with the small (less than 20 kt) GALENA and DIVIDER tests probably also occurred after the shots, although the deformation detected in these cases contains additional contributions from coseismic processes, since the radar scenes used to construct the deformation interferogram bracketed these two later events. The dimensions of the areas of subsidence resulting from all three events are too large to be solely accounted for by processes confined to the damage zone in the vicinity of the shot point or the collapse chimney. Rather, the subsidence closely corresponds to the spall dimensions predicted by Patton’s (1990) empirical relationship between spall radius and yield. This suggests that gravitational settlement of damaged rock within the spall zone is an important source of post-shot subsidence, in addition to settlement of the rubble within the collapse chimney. These observations illustrate the potential power of InSAR as a tool for Comprehensive Nuclear-Test-Ban Treaty (CTBT) monitoring and on-site inspection in that the relatively broad (~100 m to 1 km) subsidence signatures resulting from small shots detonated at normal depths of burial (or even significantly overburied) are readily detectable within large geographical areas (100 km × 100 km) under favorable observing conditions. Furthermore, the present results demonstrate the flexibility of the technique in that the two routinely gathered satellite radar images used to construct the interferogram need not necessarily capture the event itself, but can cover a time period up to several months following the shot. In order to explore the power of InSAR to discriminate underground nuclear explosions from other seismic and subsidence events, we model settlement within the chimney and spall zones as three-dimensional distributions of point volume deflation and closing tension crack sources within prescribed subsurface volumes representing the spall and collapse damage zones. We investigate the uniqueness of fitting the InSAR observations by examining the trade-offs among the spatial and strength distributions of the sources, and the overall geometries of the damage zones. Modeling constraints are based on characteristics of collapse chimneys and spall zones deduced from available in-situ observations at NTS and elsewhere, analysis and interpretation of near-source ground motion data, and existing results of dynamic modeling of contained underground explosions. We discuss refinements to the modeling approach based on analyses of the mechanics of settlement within damage and rubble zones taken from the fields of mining and geotechnical engineering, and examine the potential of the InSAR technique to detect and discriminate underground nuclear tests in different geological environments.