In step 1, data are loaded from the text file output by the relative-gravity meter (Scintrex or Burris). For each station occupation, individual readings can be deselected as needed, using the time-series plot of gravity readings as a guide. For larger surveys, stations may be grouped into loops as needed ("loop" refers to the set of station occupations for which a common drift correction is applied).
In step 2, loops are corrected for drift. Drift-correction options are:
None
Include in network adjustment
Roman menthod (interpolate)
Continuous model
Each loop can be assigned a unique drift method.
In step 3, the drift-corrected gravity differences are adjusted using the least squares method. Datum observations ("fixed" stations, i.e., those with a known gravity value) are assigned and/or loaded from a text file. Adjustment statistics are evaluated and, if needed, gravity differences can be excluded as outliers.
For all but the simplest surveys, it is usually worthwhile to complete all 3 steps before deselecting individual samples in the Data Table, or before spending much time on the drift correction. The reason is that errors in station name, tide correction, loop organization, etc. are not uncommon. It is more efficient to reveal these errors, by carrying out the network adjustment, before fine-tuning the individual occupations. For example, in the instance of station name errors, the imported data text file should be updated, and the file re-imported.
