Technical description of dgf2017a.snx
          -------------------------------------

 1.  Purpose: Datum-free VLBI session normal equation files in SINEX 2.0 
     format, including station/source coordinates and EOP.

 2.  Analysis Center: Deutsches Geodaetisches Forschungsinstitut (DGFI-TUM), 
     Technische Universitaet Muenchen, Munich, Germany

 3.  Short description:
     The sessions were selected considering the following criteria (with a
     few exceptions):
      - database version >= 4 available
      - 24 h or longer sessions only
      - at least three telescopes participating (each with more than 100 obs)
      - at least 250 observations per session
      - only non-mobile telescopes
      - only telescopes with sufficient data allowing to determine reliable 
        velocities

 4.  Estimated parameters:

     a. Celestial frame: Yes

     b. Terrestrial frame: Yes

     c. Earth orientation: Pole coordinates and their first derivatives, 
        UT1-UTC and LOD, dpsi, deps, each valid for the duration of the 
        session (normally 24 h), reference epoch is mid of session

     d. Zenith troposphere (per station): one offset for the whole session 
        plus a piece-wise linear function with rates valid for one hour,
        each of the rates constrained to 0 with 10 mm/sqrt(h) 

     e. Troposphere gradient: one offset per station for the whole session, 
        constrained to 0 with 0.5 mm, and one rate per session, constrained
        to 0 with 2.0 mm/day

     f. Station clocks: Quadratic function plus a piece-wise linear function 
        with rates valid for one hour, each of the rates constrained to 
        0 with 30 mm/sqrt(h) (clock jumps estimated additionally)
 
     g. Baseline clocks: None

     h. Other: None

 5.  Celestial reference frame:

     a. A priori source positions: ICRF2

     b. Source positions adjusted in solution: Yes

 6   Terrestrial reference frame:

     a. A priori station positions: ITRF2014

     b. A priori station velocities: ITRF2014

     c. Axis offsets: taken from official IVS-table
        http://vlbi.geod.uni-bonn.de/Analysis/Thermal/antenna-info.txt

     d. Reference epoch: session epoch

     e. Relativity scale: IERS Conventions 1992 ("consensus model")

 7.  Earth orientation:

     a. A priori precession model: MHB

     b. A priori nutation model: MHB (no FCN corrections applied)

     c. A priori x and y pole coordinates and rates, dUT1 and LOD: IERS 
        EOP-C04, linearly interpolated (before the interpolation of dUT1, 
        UT1R variations according to IERS Conventions 2003 are substracted 
        from the daily UTC 0h C04 values. After the linear interpolation, 
        the UT1R corrections are added again).

     d. A priori short period (~daily and subdaily) tidal variations in 
        x, y, dUT1: IERS Conventions 2003

     e. EOP estimation: Pole coordinates and rates, UT1-UTC and LOD, 
        dpsi, deps, each valid for the duration of the session 
        (normally 24 h), reference epoch is mid of session (see the
        SOLUTION/APRIORI block)

 8.  A priori geophysical models:

     a. Troposphere: hydrostatic component from surface pressure (in 
        general as reported in the NGS files) mapped with the VMF1
        dry mapping function, wet delay is estimated with the VMF1 
        wet mapping function. All a priori values for the gradients are 
        zero.

     b. Solid Earth tide: IERS Conventions 2003 (no non-correction of the 
        permanent tide; comment: this is not conform with the IAG 
        resolution but common practice)

     c. Ocean loading: IERS Conventions 2003, per station 11 partial tides 
        provided by H.G. Scherneck (ocean tide model: FES2004, without center of
	mass correction)

     d. Pole Tide: IERS Conventions 2003, using the linear approximation 
        of the mean pole path

     e. Atmosphere loading: -

     f. Thermal deformation: according to IVS conventions 
        http://vlbi.geod.uni-bonn.de/Analysis/Thermal/antenna-info.txt

 9.  Data type: Group delays

 10. Data editing: 5-degree elevation cut-off

 11. Data weighting: Only observations with a zero quality flag in NGS 
     cards are used. The weights determined during the correlation process 
     (read from the NGS files) are refined by station- and elevation- 
     dependent fractions. Elevation-dependent weighting is applied (Tesmer 
     and Kutterer 2004). All observations are subject to an outlier test 
     (data snooping).

 12. Software: DOGS_CS and OCCAM 6.1e, least squares method (LSM)

 13. Further information:
     The SINEX 2.0 files include datum-free normal equations only, no 
     solution. None of the station/source position parameters or EOP reported 
     in this file are subject to any constraints.