From: rothacher@ubeclu.unibe.ch (MARKUS ROTHACHER) Author: Markus Rothacher, Werner Gurtner Date: 28 Oct 1998 18:54:04 Subject: [IGEXMail-0045] Remarks on IGEX Data Processing Precedence: bulk ****************************************************************************** IGEX Electronic Mail 28-Oct-1998 18:54:04 Message No 0045 ****************************************************************************** Author: rothacher@ubeclu.unibe.ch (MARKUS ROTHACHER) Subject: Remarks on IGEX Data Processing Author: Markus Rothacher, Werner Gurtner Dear Colleagues, During the first analyses of IGEX data we detected a few things that might be of interest to IGEX participants: (1) In our pseudorange processing in the single station mode we determine station position: once per session receiver clock offsets: once per epoch GLONASS-GPS time offset: once per session using the broadcast satellite orbits and clocks. The results for the GLONASS-GPS time offsets typically look like this: *************************************************************** GLONASS/GPS TIME OFFSETS: DAY 298, 1998 *************************************************************** FILE STATION NAME RCV TIME OFFSET (NS) RMS ERROR (NS) --------------------------------------------------------------- 1 BORG 12205M004 R100 1036.915 0.796 2 CSIR 30310M001 R100 1045.006 0.860 3 DLFT 13502M006 GG24 90.755 0.736 4 DLRA 14208M002 R100 1038.114 0.809 5 GODZ Z18 88.031 0.809 6 IRKZ Z18 85.250 0.758 7 KHAB Z18 84.824 0.777 8 KR0G 10422M001 Z18 69.476 0.714 9 LINR R100 1049.974 0.831 10 MDVZ Z18 84.844 0.739 11 METZ 10503M005 Z18 84.485 0.719 12 REYZ 10202M003 Z18 98.258 0.733 13 WTZG 14201M012 R101 -713.787 0.873 14 ZIMZ 14001M005 Z18 85.734 0.755 --------------------------------------------------------------- Although there are still systematic biases in these results (broadcast orbit errors, satellite clock errors, ionospheric delays, ...) of the order of a few 10 ns, we clearly see, that the R100 receivers show differences between GLONASS and GPS time, which are about 1 microsecond larger than those of the Ashtech receivers. Currently the correction of UTC(SU) to UTC are of the order of 180 ns according to a recent Bulletin T. The broadcast corrections from GLONASS system time to UTC(SU) are of the order of -350 ns giving roughly a system time difference GLONASS-GPS of +170 ns (the corrections of GPS system time to UTC are of a few ns only). The ADDITIONAL offsets estimated with several Asthech Z18 are between zero and say +100 ns whereas the offsets estimated with R100 instruments are about 1 microsecond larger (around 1100 ns). This seem to indicate that there is a "GLONASS receiver clock" running in the R100 which constantly differs from the GPS one by 1 microsecond. ------------- Another hint that this may be the case is the remark in the Bulletin T: "Daily values of C1 are given in the following table. They are obtained as follows: the GLONASS data taken at the NMi Van Swinden Laboratorium, Delft, The Netherlands, for highest elevation, are smoothed to obtain daily values of [UTC(VSL)-GLONASS time] at 0h UTC; daily values of C1 are then derived from them using linear interpolation of [UTC-UTC(VSL)] --> A time correction of + 1285 ns is also applied in order to ensure continuity of C1 estimates on 1997, January 1 (MJD = 50449). The global uncertainty of daily C1 values is of order several hundreds of nanoseconds." Van Swinden Laboratory is also using a R100 receiver, and they apply this correction to smoothly continue the values determined by another institution (was it Leeds University? Receiver type?) before January 1st, 1997. The offset of approximately -700 ns seen at Wettzell might stem from the fact that an older receiver (R-101) is used at this location. (2) The Z18 receiver (at least the one running in Zimmerwald) generates phase observations the absolute values of which obviously increase with time. About two weeks after system start the values don't fit anymore into the RINEX format field width F14.3. Please check if your converter used to generate the RINEX observation files correctly re-adjusts the phase values and does not merely increase the format width. (3) There are still site logs missing for several IGEX sites. It would help us a lot if these site logs could be made available soon. We also noted that the information (MARKER NAME, ...) given in some of the RINEX files are not correct. Please check the header information in your RINEX files (it should be consistent with the site log information, too). Markus Rothacher & Werner Gurtner CODE Analysis Center -------------------------------------------------------------------------------- Dr. Markus Rothacher Phone : ++41-31-631-8591 Astronomical Institute Fax : ++41-31-631-3869 University of Berne E-mail : markus.rothacher@aiub.unibe.ch Sidlerstrasse 5 AFTP : ftp ubeclu.unibe.ch (130.92.6.18) CH-3012 Berne cd aiub$ftp after login Switzerland WWW : http://www.cx.unibe.ch/aiub/