Abstract Fourteen observing system simulation experiments (OSSE) wore conducted using the results from a mesoscale model on the Oxidation and Scavenging Characteristics of April Rains(OSCAR) experiment to test the accuracy of trajectory models. Our results indicate that the current synoptic network and observational frequency over North America are inadequate for accurate computation of long-range transport of episodic events. It appears that improving the Observational frequency would be more cost effective than improving the spatial resolution for the existing network. Reducing the three-dimensional air flow to two dimensions leads to a substantial amount of error for air parcel trajectories. Among the three simplifying assumptions—isobaric, isosigma, and isentropic—the isentropic model gives considerably better results than the isobaric or isosigma models, especially for the vertical transport.