Evaluation of NIMA GPS satellite ephemerides using NASA laser ranging data

This paper describes the use of recently collected Satellite Laser Range (SLR) data in assessing the accuracy of the Global Positioning System (GPS) ephemeris produced by the National Imagery and Mapping Agency (NIMA). The SLR data were collected by the National Aeronautics and Space Administration (NASA) during October-December 1996. At that time, NASA launched an intensive tracking campaign for the two GPS satellites which carry laser retro-reflectors. The satellites, Satellite Vehicle Numbers (SVN) 35 and 36, were to be tracked on a higher than normal priority level with nineteen NASA affiliated tracking stations for the duration of the campaign. There was less uniformity in the geographic coverage of the data than was hoped for. However, the campaign still afforded an excellent opportunity to assess the accuracy of the NIMA ephemerides. This was done by utilizing ephemerides provided by the international GPS Service (IGS) to form a baseline. The IGS provides GPS ephemerides that are considered to be accurate to better than ten centimeters. In addition, ephemerides from the Jet Propulsion Laboratory (JPL) was also examined, because the JPL ephemerides are used in the production of the IGS ephemerides and could provide another point in the baseline. NASA "quick-look" normal point data, the principal NASA SLR data product, were obtained from the NASA Crustal Dynamics Data Information System (CDDIS). Normal point data have better than centimeter accuracy and were used here to calculate SLR residuals from the NIMA, the JPL and the IGS ephemerides. The RMS values, for the SLR residuals over the three month study period, were seven centimeters for the IGS, eight centimeters for the JPL and twelve centimeters for the NIMA ephemerides. The IGS and the JPL ephemerides have a small negative bias for both SVN35 and SVN36, while the NIMA results are mixed, with SVN36 having the same negative bias but SVN35 having a small positive bias. One possible explanation for this bias could be a change in the satellite's center of mass, CM, which might have occurred since the launch date. Solving for the center of mass, when calculating the SLR residuals for each organization, removes the bias and gives an RMS of less than five centimeters for the IGS, less than five centimeters for the JPL and less than ten centimeters for the NIMA ephemerides.