Multi-technique comparison of atmospheric parameters at the DORIS co-location sites during CONT14. Issue 12 (15th December 2016)

• Record Type: Journal Article
• Authors: Heinkelmann, Robert; Willis, Pascal; Deng, Zhiguo; Dick, Galina; Nilsson, Tobias; Soja, Benedikt; Zus, Florian; Wickert, Jens; Schuh, Harald
• Is Part Of: Advances in space research. Volume 58:Issue 12(2016)
• Page Start: 2758
• Page End: 2773
• Publication Date: 2016-12-15
• Subjects: Atmospheric gradients -- Zenith delays -- Tropospheric ties -- DORIS -- GPS -- VLBI; Space sciences -- Periodicals; Astronautics -- Periodicals; Geophysics -- Periodicals; 500.505
• DOI: 10.1016/j.asr.2016.09.023
• Languages: English
• ISSNs: 0273-1177

Abstract

Abstract: The atmospheric parameters, zenith delays and gradients, obtained by the DORIS, GPS, VLBI, and numerical weather models, ECMWF and NCEP, are compared at five DORIS co-located sites during the 15 days of the CONT14 campaign from 2014-05-06 until 2014-05-20. Further examined are two different solutions of GPS, VLBI and NCEP: for GPS, a network solution comparable to the TIGA reprocessing analysis strategy and a precise point positioning solution, for VLBI, a least squares and a Kalman filtered and smoothed solution, and for NCEP two spatial resolutions, 0.5° and 1.0°, are tested. The different positions of the antenna reference points at co-location sites affect the atmospheric parameters and have to be considered prior to the comparison. We assess and discuss these differences, tropospheric ties, by comparing ray-traced atmospheric parameters obtained at the positions of the various antenna reference points. While ray-traced ZHD and ZWD at the co-located antennas significantly differ, the ray-traced gradients show only very small differences. Weather events can introduce larger disagreement between atmospheric parameters obtained at co-location sites. The various weather model solutions in general agree very well in providing tropospheric ties. The atmospheric parameters are compared using statistical methods, such as the mean difference and standard deviations with repect to a weighted mean value. While GPS and VLBI atmospheric parameters agree very well in general, the DORIS observations are in several cases not dense enough to achieve a comparable level of agreement. The estimated zenith delays from DORIS, however, are competitive with the other space geodetic techniques. If the DORIS observation geometry is insufficient for the estimation of an atmospheric gradient, less than three satellites observed during the definition interval, the DORIS atmospheric parameters degrade and show small quasi-periodic variations that correlate with the number of observations and in particular with the number of satellites. An increase in the DORIS constellation concerning more satellites and in general more observations is very likely to significantly improve the quality of DORIS derived atmospheric parameters. For the first time we tested a 6 h sampling of the DORIS gradients. Where the observations are sufficiently dense, the increased sampling results in an improvement of the agreement of the DORIS gradients with the other solutions.