Global air passenger transport demand is expected to increase, and there is concern that the current airport operation will not be able to cope with aircraft overcrowding. In this study, we developed a cellular automaton simulator that can model the surface traffic of the entire Tokyo International Airport in detail, including aircraft that are arriving, taxiing, parking, and departing, using actual track data. The simulator can reproduce stop-and-go aircraft taxiing based on aircraft interactions and runway rules. It can simulate the stochastic features of the surface traffic flow. To validate the developed CA simulation, the taxiing speed distribution, local delays, and taxiing times for each route were compared with the actual track data. They were in good agreement. The effects of stochastic surface traffic features, such as arrival rate, runway occupancy time, and taxiing route, on airport operations were quantitatively analyzed. This tool could lead to a better prediction of future air traffic and improve airport operations.