GFO CalVal Timing Issues

1
GFO Cal/Val Timing Issues

• John Lillibridge
• NOAA
• Laboratory for Satellite Altimetry

2
Overview of Mission Status

• Launch on February 10, 1998
• CPU resets fixed in Rev. E software November,
  1999
• GPS system failure
• Utilize operational Doppler and precision laser
  orbits
• Time tagging via ground-segment hardware June,
  1999

3
CPU Resets Data availability
4
Precision Timing Issues

• Time-tag stability uniformity
• Spacing between records 0.98 s
• Ultra-Stable-Oscillator clock drift
• Changes in clock frequency alter range
  travel-time counts
• Timing biases between Range Orbit
• msec level timing biases yield cm level errors at
  twice-per-rev

5
GFO Timing Stability
6
Geosat Timing Stability
7
Example of non-uniform spacing
8
Summary of Timing Stability

• GFO time-tagging is noisy at the 10 µsec level,
  compared to 1 µsec for Geosat (implies height
  error at the 0.1 mm level)
• GFO exhibits 10 Hz frame slips, but this only
  occurs in 0.1 of the data
• The uneven spacing of GFOs records does NOT
  imply an error in time-tagging

9
USO Height Correction - GFO
10
USO Height Correction - Geosat
11
Summary of Clock Drift

• GFOs Doppler beacon oscillator is drifting at a
  rate 10x worse than Geosat
• The USO height correction is large, but is
  implicitly corrected during SDR production
• The estimation of the USO height correction can
  suffer from large bookkeeping errors

12
Timing Biases - Laser Orbits
13
Height Variability - with orbit error
14
Height Variability - no orbit error
15
Summary of Timing Biases

• Before installation of the time-tagging unit
  timing biases were on the order of seconds
• Timing biases of 1 1 msec are seen after
  time-tagging unit installed - acceptable
• Removal of one and two cycle-per-rev orbit error
  yields expected sea level variability

16
NGDR Issues

• Byte-alignment problem in last 5 data fields
• Currently no slot allocated for model wet
  troposphere correction
• Cases of flagged sigma-H with good SSH
• 10 Hz heights not analyzed sufficiently yet
• Need to determine optimal SSB correction