The characteristics of the particle distribution, evolution and movement in a sonic jet release of carbon dioxide (CO2) from a high pressure reservoir are investigated. The motivation is to numerically model the sonic jet with particles, using the hitherto unknown initial particle distribution measured herein, and hence understand and numerically reproduce the experimentally observedparticle behaviour downstream of the Mach shock, including turbulence characteristics and level of agglomeration. We employ a Reynolds-averaged Navier-Stokes scheme with adaptive mesh refinement (AMR), combined with a Lagrangian particle tracker and particle distribution function. The model is seeded at the nozzle with the experimentally measured particle distribution and exploited to reproduce the observed characteristics of the jet. These releases are designed to be representative of a sonic CO2 release into the atmosphere and so provide data to help interpret how accidental or operational releases from the transport aspect of a carbon capture and storage chain might behave.