Abstract: The PVT(Potential Vorticity Tendency) method has already been widely used in diagnosing the mechanism of typhoon motion. The spatial resolution of the data used in those studies is usually several kilometers. In the present study,it is found that when using the PVT method to diagnose typhoon motion,the time interval of the data should match its spatial resolution owing to the limitation of the calculation. The PVT and potential vorticity distribution were calculated by using the time difference and spatial difference methods at the same time. When the spatial resolution of the model is increased to less than 1 km,some small-scale systems are simulated. Thus,the time interval of the data should be reduced to make sure it can represent the variability of these small-scale systems. As the potential vorticity caused by these small-scale systems can not directly influence the typhoon motion,two methods are recommended in this study to solve this problem. The first one is to reduce the time interval of the model output data as its spatial resolution increases. The second one is to keep the time interval unchanged and do spatial smoothing on the variables used in calculating potential voricity so as to filter out these small-scale systems.