Assessment of atmospheric pressure loading on the international GNSS REPRO1 solutions periodic signatures
Unambiguous, consistent and homogeneous GPS station coordinates are the fundamental requirement in the appropriate determination of geodetic velocities that are often used to derive geodetic and geophysical models for different applications. As for that, there have been significant efforts in the past decade to improve the modeling and parameterization of GPS solutions. Recently, the International GNSS Service (IGNSS) has generated REPRO1 solutions by reprocessing the historical GPS data from 1994 to March, 2010. REPRO1 solutions adopted the new absolute antenna phase center variations models along with most of the recent model parameters available by then and they are the first solutions to be consistently represented in one reference frame, IGS05. Based on the availability of REPRO1 solutions, this research has two objectives. The primary objective is to identify the remaining periodic signatures in the International GNSS REPRO1 solutions. These signatures are the impacts of short and long term mismodeled and unmodeled effects from both known and unknown phenomena. As a parallel activity, this research will try to explain the signatures by correlating them with different effects that have either not been modeled or modeled differently with a specific attention to the atmospheric pressure loading (APL). The secondary objective of this study is to perform the harmonic analysis investigation of weekly time series in position and residual domain of REPRO1 solution using Least Squares Spectral Analysis (LSSA) and Least Squares Coherent Analysis (LSCA) with and without APL corrections. Based on the resulting least squares spectra, the impact (benefits) of APL corrections in the present solutions have been assessed as a basis of formulating recommendations in future similar reprocessing campaigns. In order to accomplish the research objectives, a set of twenty nine (29) stations (part of the present IGNSS network) were selected in a manner which would portray the global overview. Thereafter, the selected stations analyzed using Least Squares Spectral Analysis (LSSA) and Least Squares Coherent Analysis (LSCA) frequency domain multiplications with and without the impact of APL from GGFC model. The investigations were carried out at both REPRO1 positions and residuals domains. Based on the LS spectra results, it is evident that periodic signatures are still present in the REPRO1 solutions for most of the stations under study and they appear as spectral peaks. Furthermore, the observed signatures appear to be consistent around the first to fourth draconitic harmonics with respective periods of 351.2, 175.6, 117.1 and 87.8 days, within a range of ± 14 days (±0.04 CPY). It was also observed that, there is a slight improvement to spectral peaks that may result into slight improvement of coordinate repeatability if APL were included in the processing. However, the pattern was neither clear nor consistent at different harmonic levels of the same station as well as from one station to another. Furthermore, it was also observed that, the APL does not cause any significant reduction in spectral peaks that are still present in the REPRO1 solutions. This suggest that most of the remaining signatures could be attributed to other un-modeled displacements such as non tidal loading displacement, high order ionosphere terms and mismodeling effect in GPS attitude models. To ascertain the findings, independent solutions for YELL and NRC1 were generated (1995-2010) using Bernese v5.0 software in a baseline mode, in conjunction with latest IERS models. The computed solutions were verified to be compatible with present solutions within a range of ±2.5 cm. Thereafter the computed solutions were analyzed with and without the impact of APL using LSSA and LSCA as a basis of recommendations and future work.