Generally, simulation includes all methods that can reproduce the processes of a system in an analog or digital fashion. It includes numerical simulations, and simulations quite different from numerical simulations. For example, we can simply use a circuit to simulate the generation of Earth's magnetic dynamo(one-disk and two-disk geodynamo models). In this case, we use a circuit to represent the environment of inner earth and by changing the capacitance, resistivity and the combination of the disks and electric wires to achieve different behaviors of the system. In this simulation activity, we only use several oversimplified basic components to simulate the system. Though we can use this simulation to study the geodynamic process qualitatively, it is impractical for us to use the numerical results of such simulation to draw further conclusion. Another example is flight simulation, which is widely used in flight training and entertainment. The concentration of the flight simulation is to give the person who operates the system a visual sensation of flight. In practice, we can use well organized two-dimensional objects to simulate the visual feeling of 3-dimensions. The result is that, though we are seeing 2-dimensional objects, we have the feeling that what we are seeing is 3-dimensional. Though physically this is totally wrong, it is acceptable, as the goal of flight simulation has been reached. Of course, we also can not use the result of such simulation to get any physical understanding of the system. In contrast, numerical simulation is the kind of simulation that uses numerical methods to quantitatively represent the evolution of a physical system. It pays much attention to the physical content of the simulation and emphasizes the goal that, from the numerical results of the simulation, knowledge of background processes and physical understanding of the simulation region can be obtained. In practice, numerical simulation uses the values that can best represent the real environment. The evolution of the system also strictly obeys the physical laws that govern the real physical processes in the simulation region. Then the result of such simulation can have a good representation of the real environment. From the result of such simulation we can safely draw proper conclusions and have a better understanding of the system.