The National Survey Authority (NSA) of Oman has embarked on a novel project to establish the Oman National Geodetic Datum (ONGD14). The primary objective of the ONGD is to create a homogenous horizontal survey control for the country and also provide the GPS user community with modern surveying infrastructure. The project would be carried out by strengthening the existing national geodetic control through the enhancement of the old control network and further establishment of GPS Oman National CORS Network (ONCN) stations. Although the NSA has been working in this direction for a number of years now, it was realised that a new geocentric datum based on the latest global ITRF system solution was the need of the hour in the country. The existing WGS84 (ITRF89) system had to be updated so that NSA could provide the country with the latest global and accurate geocentric datum of ITRF2008 epoch 2013.
The task began with selection of 20 existing GPS primary and first order control stations covering the entire country. The campaign covered seven primary stations and 13 first order GPS stations. A total of five GPS field teams were deployed by the NSA for the campaign with five stations being occupied per session by each team. For each of the stations, the GPS observations were carried out between seven to 23 days using five GPS receivers. Subsequently, a main network was created comprising an initial 56 selected IGS stable stations.
GPS observations were carried out for 20 NSA stations between January 26 and March 25, 2013. Data in T01 format were converted to Rinex Version 2.11. Downloading of Rinex Data from IGS Global Data Center (GDC) was carried out using customised file transfer protocol script for 59 stations. Verification of Rinex header for all stations (NSA and IGS) was done using TEQC program from UNAVCO. Marker names, receiver type, antenna type and antenna height are the main parameters for verification purposes. Downloading of IGS final orbit and IGS final earth rotation parameters (ERP) from IGS central bureau were also carried out. On-Line Ocean Loading computation from Onsala Space Observatory (OSO) was carried out for 20 NSA stations using FES2004 model. IGS final orbit was used in all computation with final IGS earth rotation parameter (ERP) from IGS central bureau. Two programmes associated with the orbit computation in Bernese 5.0 are PRETAB and ORBGEN. Ocean tides correction OT-SCRC model was introduced at this stage with development planetary ephemeris (DE200). RMS error of 1–2 cm for each satellite was achieved, which showed that the final IGS earth rotation parameter is consistent with the weekly pole information from the IGS final orbit.
Selected IGS stations
From the initial 56 planned IGS stations, seven stations had data outage for the entire observation campaign. Daily data availability for all the IGS stations was more than 50 days except for DAEJ with 46 days, ZAMB with 25 days and KUNM with 6 days. GPS observations for NSA stations were between seven to 23 days. However, due to corrupted data, the observation data for only four days is available for NSA 1066. Daily observation data for NSA stations spanned from a few hours to 24 hours. Most of the short data came from the first and last days of the observation campaign.
Combination of solutions
Normal equations in Bernese Software may be stored by the programmes GPSEST and ADDNEQ2 for a sequence of solutions, including a large number of parameter types (coordinates, troposphere, orbit parameters, etc.). The special features of normal equation stacking methods allow for extremely rapid and flexible computation of many solution types, without going back to the original observations. Normal equation from daily solutions can be combined in a multi- session solution with ADDNEQ2 program.
A total of 59 daily normal equations were divided into eight multi-session solutions with the first seven combinations comprising of 7 and the last one 10 daily solutions. The combination of solutions from DoY 026 to DoY 032 coded as Block 01 consisted of 52 stations including four NSA stations, namely NSA1058, NSA 1066, WPC3 and WPD1. Number of solutions for each station was between two to seven days with 17 station solutions eliminated due to high residuals in one of the components.
Final combination solution
The geodetic datum of the network must be defined based on some reference sites with well-known coordinates, especially for non-global networks. To ensure consistency with the orbits and earth orientation parameters, it is recommended to include some nearby stations from the global IGS network as reference sites. The precise IGS coordinates and velocities for these stations may then be used for datum definition purposes. A no-net-translation condition or tight constraints on the single reference stations is well-suited for a final coordinate solution. It is advisable to check the performance of the fiducial sites and, if necessary, to remove problematic stations from the list of reference sites. Using coordinates of one or several reference sites given in a well-defined reference frame, the estimated coordinates can be aligned to that frame. In contrast to the absolute geometry, the internal geometry of the network is very well determined by GNSS measurements because a shift of a single station in a network cannot be compensated by simply adjusting
- Combination of eight multi-session normal equations was achieved.
- 48 selected IGS stations were held fixed with the coordinates transformed to an epoch of the middle of the campaign
- Free network adjustment with the introduction of 3-parameter Helmert’s transformation was applied to the campaign solution
- Results were analysed statistically for coordinate repeatability and RMS of residuals
- Bad solutions were excluded at this stage
- The RMS of campaign solution noted to be less than 10 mm in the horizontal and height components
Results and analysis
The final combined solution consists of eight multi-session solutions with 68 stations (20 NSA stations and 48 IGS stations). Minimal constraint adjustment using free network solution with 3-parameter Helmert’s transformation was used to adjust the daily normal equation freely and transform them using initial 48 selected IGS stations. With short data span, the introduction of reference velocity for the fixed stations is not possible; hence, the final coordinates for all stations were fixed at the middle of the observation epoch. RMS of residuals was 2.00 mm, 2.90 mm, and 4.26 mm for northing, easting and height components respectively. It can be concluded that the internal accuracy of all stations from the free network adjustment is less than 5 mm in all components.
Comparison of IGS station coordinates was done in order to determine the accuracy of the network with respect to the IGb08 reference frame. The RMS of fitting came out at 3.9 mm, 5.5 mm and 8.3 mm for the northing, easting and height components, respectively. It can be concluded that the accuracy of NSA stations with respect to the IGb08 reference frame with free network strategy is 5.9 mm to 8.4 mm in the horizontal component and 12.56 mm in height.
The revision of a geocentric datum is inevitable considering that satellite positioning systems would have widespread use in this millennium and the positions referenced to the existing datum would not be compatible with the updated satellite-derived positions. The revision of a global geocentric datum based on the current reference frame would allow the implementation of Network Based Real-Time GNSS Services (RTKNET) for a single standard for the acquisition, storage and the use of geographic data, thus ensuring compatibility across various GIS applications.