Time-series analysis is a key technique in interferometric synthetic aperture radar (InSAR), which is widely used to monitor slow surface deformation over large areas. This method can provide the ability to monitor deformation over a wide range and large area. Among them, the StaMPS PS method has been favored by many scholars due to its advantages of wide application range and open source. However, in time-series InSAR, the delayed phase of the troposphere will lead to the degradation of the deformation monitoring accuracy. Therefore, this paper analyzes the linear correction of empirical model, the GACOS (generic atmospheric correction online service for InSAR) correction and the latest release of the European Center for Medium-Range Weather Forecasts ( ECMWF) with the city of Hefei as the study area. ECMWF (European Center for Medium-Range Weather Forecasts), and compare the correction effects of these three methods in the time-series InSAR inversion of deformation rate. The standard deviations in the study area are obtained by calculation, and the related reasons are analyzed and validated for comparison. The standard deviations of linear correction, ERA5 and GACOS correction are reduced by 21.71%, 16.14% and 10.38%, respectively. For the Hefei region, these three methods can attenuate the effect of tropospheric delay and improve the accuracy, among which the linear correction has the best effect and higher applicability, while the ERA5 and GACOS are affected by the weather and the density of the ground monitoring points, which have poorer correction effect in this region.