由于越来越多研究需要从观测或者再分析资料中分离台风环流和环境气流，尝试将多尺度窗口变换（multiscalewindow transform，MWT）方法应用于分离台风环流，并与空间分离法和Lanczos滤波法进行了比较。理想化试验表明，MWT方法在用于对台风环流的分离时，比另外两种方法误差小。当再分析资料中台风强度较强时，单一分离方法都存在较大的误差。为了减小台风环流分离中的误差，可以先利用空间分离法初步滤去台风环流，然后再利用MWT方法或Lanczos滤波法进一步分离台风环流。

More and more researches need to separate tropical cyclone circulation from its environment in the observation and reanalysis data. The multiscale window transform (MWT) method is applied to the separation of tropical cyclone circulation from its environment and compared with tropical cyclone spatial removal technique and Lanczos filter in time. Through using these approaches to idealized wind fields, the MWT method is better than the other two approaches in the separation of tropical cyclone circulation. It is also found that the performance of the combined approach is better than that of individual approaches. In order to obtain better results, the spatial removal technique can be used first to separate tropical cyclone circulation and then the MWT or Lanczos filters in time is used for further separation of tropical cyclone circulation.