Modeling atmospheric effects on InSAR with meteorological and continuous GPS observations: Algorithms and some test results

A model for the integrated use of continuous GPS (CGPS) and ground meteorological observations for modeling and correcting atmospheric effects on InSAR measurements is presented. The model takes into account the spatial relationships of the two types of data and the elevation-dependent nature of the atmospheric delays. It overcomes the shortcomings of some existing methods that model only the homogeneous part of the atmospheric effects, and reduces to some of the existing models under certain conditions, e.g., when the area of interest is too far from the sampled locations. Tests with some datasets in Hong Kong (HK) show that the RMS values of the total zenith delay (TZD) models developed based on CGPS observations only and on both CGPS and ground meteorological observations are around 0.5 cm, while that based on ground meteorological observations only is about 4 cm, although the relative accuracy of the model is much higher. The HK dataset shows that the temporal and spatial variations of the tropospheric delays can potentially cause a peak-to-peak path error of about 7.8 and 9.5 cm in a SAR interferogram at the 95% confidence level for the 1- and 10-day intervals, respectively. When corrections are applied, the peak-to-peak errors are reduced to 6.6, 5.5 and 4.9 cm, respectively, for a 1-day interval for the three different models used, and 7.9, 6.2 and 5.8 cm, respectively, for a 10-day interval and the three models. The accuracy improvements for an interferogram would range from 15% to 39% when corrections are applied based on the models.