This problem can occur when you import an RTK job where the reference triplet is slightly different than the existing reference triplet of that station. It can also occur in Drag and Drop operations from post-processed jobs where a reference triplet has slightly different coordinates. Unfortunately, as you've probably already found out, you cannot simply change the name of the offending duplicate point.

You have two choices to prevent this from happening and a third to correct it after it happens.

1) With SKI-Pro 2.1 and above, you can modify the reference triplet of the point to be imported so that it agrees with the existing reference triplet (or vice versa) before assigning the import.

2) You can export the separate projects to SKI ASCII baseline files, and import them into a new project.

3) You can export the existing project as SKI ASCII baseline files, edit the ASCII file by removing the (2)'s in Overwrite (not Insert) mode (be careful not to move the columns in the file), and finally import the edit baselines into a new project.

This problem occurs when you use partial control, i.e. you fix only the horizontal position or only the height of a control point, and export using Main or Current as the coordinate class. The Main or Current class will export the points used as control from the Control class, and the other points from the Adjusted class. The problem is that when you only fix the height of a control point, the horizontal position is not updated in the Control class. It is only updated in the Adjusted class along with the fixed height. When you only fix the horizontal position of a point, the height is not updated in the Control class. Exporting with Main or Current as the class works fine when your control is all fixed in 3-Dimensions, but not when you only fix the height, or the horizontal position.

* When using any partial control, always select the coordinate class as Adjusted for exporting, or for printing from the Points tab listing.

When you import data and process it, it is automatically part of the next adjustment. You can deactivate the existing vectors from the day before and then adjust only the new vectors. If you do not want the points that are common in the two days to be adjusted again, fix the common points as control and they will not be readjusted.

However, there really is no reason to bother with this. If you are just doing free adjustments while building a network, simply add the new vectors and readjust each day. When you have a final free adjustment, free of blunders, fix any control values for the constrained adjustment. Remember, if you use partial control, i.e. vertical only or horizontal only, export the Adjusted point class, not the Current or Main class.

When an adjustment is stored (manually or automatically) the hidden points are recalculated to the adjusted auxiliary points. If the auxiliary points are changed any other way, you have to manually update the hidden point positions from the "Compute Hidden Points" option of the "Tools" menu.

The weight model is selected in the configuration screens reached by right-clicking in the Adjustment screen. See the settings in the examples below:

SKI-Pro - Adjustment A Priori Weighting Selections

Long and short baselines (using base error plus ppm weight model)

Configure - General Parameters

Standard Deviations

             GPS Baselines   0.005m     1ppm (change if not realistic)

             Compute Using   The default settings for all observations

                                           (software to use above weight values)

Test Criteria

            Sigma a priori  1.0 (no scaling, estimates are realistic)

Balanced baselines (using processing variance/covariance weight model)

Configure - General Parameters

Standard Deviations

            GPS Baselines   0.005m     1ppm (ignored in this case)

            Compute Using   Individual settings for all observations

                                          (software to use processing weighting)

Test Criteria

            Sigma a priori  10.0 (optimistic statistics scaled by 10)

 If the constrained adjustment fails the F-Test, multiply the current value of the Sigma a priori by the square root of the F-Test results and compute again.

Thanks to Joseph McNichols of Norfolk County, Mass. for the impetus on this.