We have performed a computational study of the experiments performed by Lowry et al. at the Arnold Engineering Development Center. In these experiments, an rf discharge is used to weakly ionize a volume of air; then a projectile is fired through this plasma. Relative to the conditions without the discharge, the shock standoff distance is observed to increase substantially, and the bow shock becomes flatter. We have modeled the rf discharge and the resulting thermochemical state of the air within the discharge region. Based on these conditions, the projectile flowfield was simulated to determine whether the relaxation of the stored internal energy causes the observed shock movement. The results indicate that the stored internal energy does not relax fast enough to reproduce the experimental results, and, therefore, vibrational energy storage is not responsible for the observed shock movement. We consider two additional mechanisms to explain the experiments: modification of the electric field by the presence of the metallic projectile, and thermal nonuniformities in the plasma. The latter effect appears to provide the best explanation for the observations. We have also modeled experiments in which microwave-discharge excited air flows over a model. Unsteady thermal effects in the pulsed discharge can account for most of the observed drag change.