During the past several years the lower stratospheric winter-season conditions, including cold temperature and a strong polar vortex, have persisted and extended to the normal spring season. Based on nearly 20 years (1979-1998) of NCEP/NCAR reanalysis data, we use potential vorticity (PV) area diagnostics to study the long-term variation of the polar vortex associated with the long-term trends of stratospheric ozone and temperature and the tropospheric wave forcing. We find that in the southern hemisphere (SH) the polar vortex has lasted about 2-3 weeks longer in the 1991-97 period than in the 1979-84 period and the northern hemisphere (NH) polar vortex has lasted 3-4 weeks longer. Further examination indicates that the SH vortex persisted within the layer (12-22 km) with almost complete ozone loss characterized by the Antarctic ozone hole, but did not persist in the higher altitudes where ozone was not depleted. However, the NH vortex persisted in a broader vertical range not limited to the ozone-depletion layer. The SH springtime photochemical ozone destruction seems to be the main cause in which enhanced radiative cooling stabilized the polar vortex. In the NH the process is more likely due to the decadal stratospheric cooling trend and changes in the general circulation, which is related to natural and anthropogenic activities . To verify this, we examine and compare such dynamical parameters as polar temperature, the E-P flux and wave potential energy, in both hemispheres. We find that in recent years there have been more dramatic changes in those dynamical parameters in the NH than in the SH. These results indicate the complexity of events in the two hemispheres and that we must consider both the radiative and dynamic interaction effects of long term changes