Statistical inference in the presence of an inclination effect in laboratory radiative shock experiments

A laser-driven experiment produces images of dense shocked material by x-ray transmission. The post-shock material is sufficiently dense that no significant signal passes through the dense layer, and therefore the shock compression cannot be directly measured by comparing transmitted intensities. One could try to determine the shock compression ratio by observing the ratio of the total distance travelled by the shock to the dense post-shock layer width, but small deviations of the angle of the shock with respect to the angle of imaging create large asymmetric errors in observation. A statistical approach to recovering shock compression by appropriately combining data from several experiments is developed, using fits to a simple model for the shock and shock tube geometry.