Regarding the 8888 tie file:
  • A bump occurred at around 15:55, and the gravimeter was brought back
    to the previous station occupied. I appears like a tear occurred,
    increasing the readings by 30μGal after the event. To account for this
    measured effect, a virtual pair of check-in was created to tightly 
    constrain the event, one at 15:55:00 (5937.7759 mGal), the other at
    15:55:01 (5937.8059). These two values were calculated to equalize the
    slopes going respectively from morning check-in to just before the tear, 
    and from the tear to the evening check-in.

Regarding the 3333 tie file:
  • A virtual check-in was created for the end of the day, because another
    gravity check-in point was used and 3333 wasn't revisited. To do this:
      – A Δ from datum value was calculated for the check-in used at the 
        begining of the day:
          ○ station 2222
          ○ datum                          7377.834 mGal
          ○ early  day value             – 6001.348 mGal
          ○ Δ                            = 1376.486 mGal.
      – The Δ value was then subtracted from the datum of 3333 to create 
        a virtual check-in at the time of the 2222 check-in:
          ○ station 3333
          ○ datum                          7335.710 mGal
          ○ Δ                            – 1376.486 mGal        
          ○ virtual check-in             = 5959.224 mGal
  • Following this first step, the effect of the tear described in the 
    first section (above) was included in the same manner as for the 
    8888 tie file, similarly resulting in the addition of a pair of virtual
    check-ins at 15:55:00 (5959.197 mGal) and 15:55:01 (5959.227 mGal).
    
Regarding the 2 xyz files:
  • The DEM covering the whole area is very large, in fact likely too large for
    the field computer used in conjunction with Gravred 3.2 (who is systematically
    loading the whole DEM to the RAM before proceeding to calculations). So in order
    to avoid errors due to lack of memory available, the inner DEM was split into
    2 files of 5.5M points each, and the results from two individual runs are added
    together to create the final xyz containing one unique inner DEM correction value.
  • The final xyz is actually a merger of all days together.