Static GNSS precise point positioning using free online services for Africa

The GNSS precise point positioning (PPP) technique has been extensively covered in recent research. In this study, six International global navigation satellite system(GNSS) Service (IGS) stations defined by 4-digits code in Africa were selected to be processed. The stations cover different climates in Africa. HARB, SUTH and WIND stations cover the mid-latitude area and NKLG, NURK and MAL2 are located in the equatorial region. Two data sessions were selected in processing to cover the winter and summer seasons: the first session of 3 days [01-03 January 2013 (DOY: 001-003/2013)] and the second session of 3 days [01-03 July 2013 (DOY: 182-184/2013)]. This paper aims to evaluate the accuracy of static PPP coordinate solution in Africa for different convergence times using free online services: Canadian Spatial Reference System (CSRS)-PPP, Automatic Precise Point Service (APPS)-PPP and GPS Analysis and Positioning Software (GAPS)-PPP. The stations' observation times were divided into different observation times (1, 2, 4, 8, 12, and 24 h). The PPP coordinate solution was compared to the reference solution of those stations. The investigation is continued by estimating the PPP total tropospheric zenith delay (TZD) parameters, which were obtained by the different online services. These parameters were compared to the published parameters from IGS. The comparison between the different online services shows that the CSRS-PPP provides the best solution after 4 and 8 h. The APPS-PPP and CSRS-PPP provides the same solution in millimeters level after 12 h. In the equatorial region, the PPP coordinate solution is significantly improved between 12 and 24 h. Moreover, GAPS-PPP is not recommended to be used in the equatorial stations, where it shows a high error compared to CSRS-PPP and APPS-PPP even after 24 h. The mid-latitude stations show a better PPP coordinate solution in the winter than in the summer, but the equatorial stations present a low accuracy for the two sessions. Regarding the PPP tropospheric delay estimation, an additional station (HNUS) is used in the mid-latitude area, which has a low ellipsoidal height. APPS-PPP shows the best solution in the TZD estimation. The GAPS-PPP online service shows a systemic error in estimation and a high RMS relative to the TZD values known from IGS. The ellipsoidal height for antenna indicates a negative correlation to the estimated tropospheric values, but there is no effect for the variation of the ellipsoidal height in the estimated RMS. The estimated tropospheric parameters are correlated with the PPP coordinate estimation, where the GAPS-PPP service provides the worst PPP coordinate solution and at the same time, it shows the worst accuracy for tropospheric parameters estimation. Moreover, the estimated tropospheric parameters for the mid-latitude stations in the summer season show a higher RMS than in the winter season, which matches the same sequence of the PPP coordinates. The NKLG and NURK stations present a low accuracy for the tropospheric estimation, which is matching to the low accuracy of the PPP coordinates obtained from the different online services.