Estimation of the precision of atmospheric phase delay models for displacement fields on Kamchatka region calculated by the differential interferometry method

The paper considers various ways of calculating atmospheric phase delays to correct the interferometric phase from which the displacements of the Earth’s surface are calculated by satellite radar interferometry methods. A local empirical weather model based on the empirical dependences of the physical properties of the atmosphere on altitude was constructed using data from Kamchatka meteorological stations, and electromagnetic signal propagation delays were calculated using this model. Further, the precision of the empirical weather model and the GACOS (Generic Atmospheric Corrections Online Service) model was assessed by comparing them with the delays calculated from the network of GNSS sites of the Kamchatka Branch of the Federal Research Center of “United Geophysical Service of RAS” (KB of UGS RAS) in the area of the Kliuchevskoi group of volcanoes. The results showed that for all precision assessment criteria, the GACOS model has lower error and better match with the GNSS data at the result comparison points. The relative residuals of the empirical model delays range from 0 to 5.7%, while for the GACOS delays the relative residuals do not exceed 1.6%. Meanwhile, for the GACOS and empirical weather models, on average, the relative residual is 0.3% and 0.9%, respectively, and the RMS errors are 0.6 cm and 2.3 cm. In general, at points of the locations of GNSS sites, a fairly good precision of calculations was obtained: the error is less than 1%, as well as a very high coefficient of determination of the dependencies of the compared models, almost corresponding to the correlation coefficient equal to 1. In addition, for 25% of the results, it is obtained that the empirical model outperforms the GACOS model, i.e., the delay values are closer to the delays calculated from the measurements at the GNSS sites. Since half of the weather stations and GNSS sites used to calculate the empirical weather model are local stations of the Kamchatka network that are not part of the global networks, the results obtained provide an independent assessment of the precision of corrections from the GACOS online service to the peninsula. The results of the study also show that the empirical atmospheric model constructed in this work provides good precision of calculations at GNSS point locations and their data allow us to calculate atmospheric delays for the Earth surface displacement fields obtained by satellite interferometry methods with high accuracy.

satellite interferometry, synthetic aperture radar (SAR), Kamchatka, GACOS, atmospheric delay, weather model