Atmospheric delay is one of the most significant errors in InSAR. However， efficient methods have not been found to correct such artifact yet. Weather models have been widely used to correct InSAR atmospheric delay errors in recent years

but more tests are required to demonstrate the efficiency of this method. The ECMWF model was used to correct InSAR atmospheric delay errors in central Tibet. The results indicate that only 3 out of 8 interferograms have been improved with smaller variance after correction

and the average variance is even bigger than that before correction. It means that with the ECMWF model

reliable improvements cannot be made in our study area. Considering that there is no reliable approach to test when and where the ECMWF model can improve the quality of InSAR data

it is suggested that the users should be more careful about using this model for InSAR correction.