The images are Energy Time Curve (ETC) plots for the same source and the same receive location. First the measured data, then the data from the room simulation.
In cases (like the ETC measured v. modeled example above) where this pulse is "false" data we have means to remove it from our model. If it is too high in level, we assume that either the absorption or scattering coefficients need to be adjusted. If we built what most would consider a geometrically correct model with acoustically "correct" surface materials, the answer is found by adding scattering.