GPS and GLONASS observable-specific code bias estimation: comparison of solutions from the IGS and MGEX networks

• Published in: Journal of geodesy 2020-08, Vol.94 (8), Article 74
• Main Authors: Wang, Ningbo, Li, Zishen, Duan, Bingbing, Hugentobler, Urs, Wang, Liang
• Format: Article
• Language: English
• Subjects: Bias; Computation; Earth and Environmental Science; Earth Sciences; Geodetics; Geophysics/Geodesy; Global positioning systems; GPS; Ionospheric models; Original Article; Parameterization; Satellites; Stability
• ISSN: 0949-7714; 1432-1394
• DOI: 10.1007/s00190-020-01404-5

Different from differential code biases, the observable-specific code biases (OSBs) directly describe the biases of individual pseudorange measurements, which provide full flexibilities for multi-GNSS code biases handling. We present the method for the parameterization, computation and alignment of multi-GNSS OSBs as part of the local ionospheric modeling. As a representative example, GPS L1/L2/L5 and GLONASS L1/L2 OSBs were estimated during 2017–2018 from the independent International GNSS Service (IGS) and its multi-GNSS experimental (MGEX) network stations. The stability of the estimated satellite OSBs is at the level of 0.06–0.12 and 0.09–0.15 ns for GPS and GLONASS, respectively. The bias root-mean-square (RMS) differences between IGS- and MGEX-based OSBs generated by the identical estimation method are on the order of 0.1–0.2 ns for GPS and two times worse for GLONASS. The comparison between GPS L1/L2 satellite OSBs during August and October of the years 2014 and 2017 reveals that the stability of OSB estimates during high solar conditions is around 1.5 times worse than that during low solar conditions for both IGS and MGEX solutions. To check the sensitivity of OSB estimation results to distinct receiver types, the bias discrepancy between different groups of receivers (i.e., Javad, Septentrio and Trimble) was investigated during a 3-month period in 2018. The maximum OSB RMS difference between different groups of receivers is found to be 0.6–0.9 ns for GPS and 1.4–1.7 ns for GLONASS. The reason might be that the response of receivers’ correlator and front-end designs differs between receiver manufacturers. As such, further investigation has to be carried to take into account the different groups of receivers, i.e., groups of several receiver models/brands which exhibit similar OSBs.