Modeling and application of the time-varying GPS differential code bias between C1 and P1 observations

Abstract

Considering the time-varying characteristic of the GPS differential code bias between C1 and P1 observations (DCB (C1-P1)), the method for estimating the time-varying DCB (C1-P1) is presented. The time-varying characteristic of the DCB (C1-P1) is analyzed using the estimated values with the twelve-month (1 Jan-31 Dec 2018) GPS data set from 120 International GNSS Service (IGS) stations. The estimated results show that the DCB (C1-P1) has the evident variation with the periods of 12, 8, 6, 4, 3.4286 h. Based on the noticed periods, the estimated single-day DCB (C1-P1) series is modeled with a harmonics-based function. The results show that the initial phase offsets (θi(i=1~5)) have the characterization of linear variation and these variations are very evident, while the amplitudes (yi(i=1~5)) of each satellite irregularly vary with time. The time-varying DCB (C1-P1) is used in single point positioning (SPP) to validate its effect on positioning. The SPP results show that the three-dimensional positioning results can be improved, when the time-varying part of the DCB (C1-P1) is corrected. The results for the 120 IGS stations indicate that the average three-dimensional positioning improvement for three days is about 3 cm. These improvements further advise that the time-varying part of DCB (C1-P1) should be corrected in GPS data processing and be serviced by IGS. The prediction performance shows that the correction rates of the predicted results for all satellites reach more than 40% accuracy and the mean prediction accuracy can reach 70%.