REALTIME TIME BIAS COMPUTATION AND DISTRIBUTION *********************************************** Werner Gurtner March 26, 2002 The following proposal is a result of discussions at the recent EUROLAS workshop in Herstmonceux. TIME BIAS FUNCTIONS The NERC Space Geodesy Facility group at Herstmonceux is automatically and frequently collecting normal point data to compute updated time bias functions (with respect to all available IRV sets) for all ILRS satellites. FASTER DISTRIBUTION In order to simplify, and greatly speed up, the distribution of these updated time biases, I have written a TCP/IP server program that accesses the latest time bias functions at NERC and computes time biases for the current epoch, including drag functions (if existing) for all available satellites and IRV sets. So with one simple call stations can get the best current estimates of time bias for all satellites. USAGE To download the values use the EUROSTAT program (the program used to access the EUROLAS status exchange - see SLRMAIL 372) as follows: eurostat aiuli3.unibe.ch 7840 1 [>time_bias_file] The program connects to port 7840, receives the time bias values and writes them to the standard output. You can use telnet, too: telnet aiuli3.unibe.ch 7840 [>time_bias_file] On UNIX systems you can also redirect the telnet standard output to a file. Or to display the values from a web browser: telnet://aiuli3.unibe.ch:7840 or http://aiuli3.unibe.ch:8000/cgi-bin/cgi-time_bias Or you write your own program using socket library routines to connect to port 7840 and to receive the time bias message. The host disconnects the tcp/ip connection immediately after having sent the message. TIMING Depending on the time needed for range data post-processing at the stations, data submission, data handling at the data centers, and access by the analysis center, the distributed time biases could be VERY up-to-date. FEEDBACK FROM STATIONS For very low satellites (currently Champ, GraceA, GraceB) GFZ has asked stations to send normal points directly to GFZ as soon as possible after the pass: any stations doing this are requested to add Herstmonceux to the mailing list (slr@slrb.rgo.ac.uk) using the same subject header as requested by GFZ. EXAMPLE ! ! Time biases at 26-Mar-2002 11:11 UT ! ! Satellite Total IRVset LstObs Passes SIC Drag Dyn ! TB[ms] [hhh:mm] used [ms] Prio ! Ajisai 0 HON084 16:57 5 1500 0 -1 Ajisai -233 RGO095 16:57 58 1500 0 -1 BeaconC 0 HON084 0:00 0 0317 0 0 BeaconC 217 RGO040 35:26 39 0317 0 0 Champ -100 GFZ310 0:00 0 8002 -100 +1 ERS2 0 GFZ394 0:00 0 6178 0 0 Etalon1 -32 HON084 34:57 1 0525 0 0 Etalon1 10 RGO020 34:57 31 0525 0 0 Etalon2 0 HON084 0:00 0 4146 0 0 Etalon2 -6 RGO020 40:14 29 4146 0 0 GFO1 1 HON084 30:52 1 8501 0 0 GFO1 42 RGO099 30:52 28 8501 0 0 ... ... ! --------------------------------------------------- The message consists of - A number of header records - one line per time bias - last line: dashed line. The lines are separated by CR/LF pairs. All non-data lines start with "!". 1st column: Satellite name 2th column: Current time bias, drag function value included (ms) 3rd column: IRV set 4th column: Age of latest np data used for time bias function computation 5th column: Number of normal points used to generate the time bias update 6nd column: SIC code 7th column: Value of drag function (ms) (ALREADY INCLUDED in TB of col 2) 8th column: Dynamic priority Format: (2X,A10,1X,I6,1X,A6,1X,I3,1X,I2.2,1X,I5,1X,I4,1X,I4,I3) NOTE: For program-controlled evaluation of the time bias message please skip any line within the message starting with "!". The final one (dashes) indicates the end of the message. SUGGESTIONS ON HOW TO USE THE TIME BIAS SERVER A) - Compute az/el/range elements immediately before each pass using the current time bias from the time bias server B) - Compute az/el/range elements any time before the pass - either computing the predicted time bias using the time bias function (and drag function) - or downloading the time bias from the time bias server (valid at the time of computation) - Immediately before the pass: Download the current time bias and correct the initial tracking by the difference "actual time bias minus time bias used for the computation of the elements" BACKUP In order to be prepared for communication problems with the server or any other unavailability of time biases it might be a good idea to have the computation of the time bias using the distributed time bias functions (standard time bias file) ready as a backup. FUTURE The NERC colleagues are currently investigating the possibilities to include also time biases relative to the preceeding ("yesterdailies") IRV sets to cover the time around IRV updates. PROGRAM SOURCE AND EXECUTABLES The C source (eurostat.c), the PC executable (eurostat.exe_doc) and explanations to the eurostat program (eurostat.txt) can be found on ftp://ftp.unibe.ch/aiub/slr Please contact me if you have any problems or suggestions. ACKNOWLEDGMENTS I would very much like to thank Roger Wood, Philip Gibbs, and Graham Appleby of the NERC Space Geodesy Facility and Georg Kirchner, Graz for their support and many good suggestions. Best regards, Werner Gurtner