PDS_VERSION_ID = PDS3 RECORD_TYPE = FIXED_LENGTH RECORD_BYTES = 80 OBJECT = TEXT PUBLICATION_DATE = 1999-02-04 NOTE = "Description of GEODYN SPK Files" END_OBJECT = TEXT END GEODYN SPK files (data type GDN) are QUICK LOOK orbits produced by the MOLA POD (Precise Orbit Determination) team. They are used for preliminary analysis of the MOLA data, and have been made available to other MGS project participants. They are produced on a best efforts basis, as time and resources permit. Each file is in SPK format, compatible with readers in the SPICE toolkit (see companion files TK_FTP.TXT, TK_DESC.TXT, and TK_INSTL.TXT in this DOCUMENT directory for more information on the SPICE toolkit and files). The data were orginally written out by GEODYN (the orbit processor) at five or ten second intervals. They were converted from the original true of date (IAU of date) coordinates to J2000, and then were written in an SPK-format (type 8, discrete states, evenly spaced, Lagrange Interpolation). A TOPEX-like/POE (ASCII) format is available upon request from the MOLA POD Team. File names have the format gdn_qlnn.ymmddhhmm.ymmddhhmm.tsp where the components refer to: gdn derived from analysis using the GEODYN orbit processor; ql quick-look; nn version number; yymmddhhmm start and stop times tsp suffix designating an SPK transfer file Files can also be obtained via anonymous ftp from geodesy.gsfc.nasa.gov (node number 128.183.201.91). Change directories (Unix command "cd") to directory /dist/MGS/quick-look-orbits/ and retrieve the data of interest. Version 02 orbits are produced from a preliminary analysis of the Doppler and range data supplied by the DSN. They are preliminary, because certain ancillary information may be missing at the time these orbits are reconstructed, for instance the dates and times of momentum dumps, weather data (used for Earth troposphere refraction corrections), finalized Earth orientation (polar motion) values. The a priori modelling includes the GMM-1 gravity field, the 1991 IAU model, the Stewart (1987) drag model, and the DE403 ephemeris. Each arc adjusts the state vector, a drag coefficient, pass-by-pass range biases, and empirical parameters at regular intervals to absorb orbit error and other mismodeling. Version 04 orbits are derived from a reanalysis of the tracking data using an improved gravity model to 70x70 (MGM0827E) that includes the MGS tracking data from SPO-1, orbit insertion, and the hiatus (october, november 1997). Other than the improvement in the gravity model, the parameterization remains the same as the Version 02 orbits. The orbits will be refined on a schedule yet to be determined. Greg Neumann of the MOLA team has evaluated the orbits using the MOLA altimetry. For the MGM0827E orbits, the standard deviation of the crossovers in SPO-1 is 60 m; Using the square-root of two rule of thumb, the radial orbit error for the SPO-1 orbits would then be about 42 to 43 meters. For SPO-2, the crossover standard deviation is higher (168 m or a radial orbit error of 119 m via the square-root of two rule of thumb), reflecting the fact that the apriori gravity model (mgm0827e) does not yet include SPO-2 tracking. Hiatus vs SPO-1 crossovers are 60 m in the standard deviation. Version 06 orbits are released beginning on November 2, 1998. ============================================================= These orbits are based on an updated preliminary gravity model, mgm0883, a 70x70 model that includes the historical tracking (mariner 9, viking orbiter), and the tracking of MGS for both SPO-1 and SPO-2. No other changes to the processing were made, other than updating the gravity model. Version 09 orbits are released beginning on February 4, 1999 ============================================================ These orbits are based on an updated gravity model, mgm0890, a 70x70 model. The changes from the previous model, mgm0883, are the following: (1) The remaining SPO-2 tracking data from Sept. 1, to Sept. 17 1998 were included; (2) All the one-way Doppler data in SPO-1 and SPO-2 were rigorously reanalyzed and scrutinized on a pass-by-pass basis, and the Low Gain Antenna (LGA) measurement offset correction was systematically applied (in conjunction with the spacecraft quaternions which orient the spacecraft as a function of time) in the reduction of these data. Previously, in the mgm0827e and mgm0883 solutions the one-way Doppler data had not included the measurement offset correction; (3) The one-way Doppler data were upweighted in the OD solutions from 1 Hz (3.57 mm/s), to 0.2 Hz (0.71 mm/s). Greg Neumann has evaluated the SPO-1 and SPO-2 altimeter crossovers using the gdn_ql06* orbits, and the new gdn_ql09* orbits by computing the L2 and L2 norms. Remember, so far the altimeter data are an independent data type. The new mgm0890 orbits show a major improvement, especially for SPO-2. His report follows: ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Wed, 3 Feb 1999 19:05:19 -0500 (EST) From: Greg Neumann To: flemoine@ares.gsfc.nasa.gov The crossovers for SPO-1/SPO-1 and SPO2/SPO-2 show slight improvement in SPO-1 and major improvement in SPO-2. In both cases, attitude kernels were set to Dave's preferred values. Observation timing bias was set to 115.2 ms. I see a small component of drift in both groups, this bias could be upped to 116 or 117 but it is not a major factor. orbits: gdn_ql06 (0885) gdn_ql09 (0890) --------------- --------------- SPO-1/SPO-1 L2 63.94 61.57 L1 10.14 9.50 N 1188 1189 SPO-2/SPO-2 L2 141.86 70.87 L1 38.83 13.86 N 5879 5875 (slight difference due to in-plane changes) +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Direct questions to F. G. Lemoine (flemoine@olympus.gsfc.nasa.gov) or D.E. Smith (dsmith@tharsis.gsfc.nasa.gov)