Recent measurements made with the Sondrestrom incoherent scatter radar have indicated that the formation of polar cap patches can be closely associated with the flow of a large plasma jet. In this paper, we report the results of a numerical study to investigate the role of plasma jets on patch formation, to determine the temporal evolution of the density structure, and to assess the importance of O+ loss rate and transport mechanisms. We have used a time-dependent model of the high-latitude F region ionosphere and model inputs guided by data collected by radar and ground-based magnetometers. We have studied several different scenarios of patch formation. Rather than mix the effects of a complex of variations that could occur during a transient event, we limit ourselves here to simulations of three types to focus on a few key elements. The first attempt employed a Heelis-type pattern to represent the global convection and two stationary vortices to characterize the localized velocity structure. No discrete isolated patches were evident in this simulation. The second modeling study allowed the vortices to travel according to the background convection. Discrete density patches were seen in the polar cap for this case. The third case involved the use of a Heppner and Maynard pattern of polar cap potential. Like the second case, patches were seen only when traveling vortices were used in the simulation. The shapes of the patches in the two cases of moving vortices were defined by the geometrical aspect of the vortices, i.e. elliptical vortices generated elongated patches. When we “artificially” removed the Joule frictional heating, and hence any enhanced O+ loss rate, it was found that transport of low density plasma from earlier local times can contribute to ∼60% of the depletion. We also found that patches can be created only when the vortices are located in a narrow local time sector, between 1000 and 1200 LT and at latitudes close to the tongue of ionization.