Statistics of Empirical Removal of Tides and Inverse Barometer Effect on DInSAR From Double DInSAR and a Regional Climate Model